Sunteți pe pagina 1din 34

Society for American Archaeology

Conceptual Metaphor in the Archaeological Record: Methods and an Example from the American Southwest Author(s): Scott G. Ortman Source: American Antiquity, Vol. 65, No. 4 (Oct., 2000), pp. 613-645 Published by: Society for American Archaeology Stable URL: http://www.jstor.org/stable/2694419 . Accessed: 04/03/2014 07:23
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Society for American Archaeology is collaborating with JSTOR to digitize, preserve and extend access to American Antiquity.

http://www.jstor.org

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

CONCEPTUAL METAPHOR IN THE ARCHAEOLOGICAL RECORD: METHODS AND AN EXAMPLE FROM THE AMERICAN SOUTHWEST
Scott G. Ortman

This paper attempts to unify recent theorizing on cultutr-ail m7eaning in material culture using the notion of conceptual metaphor

Researchin several disciplines suggests that conventionalmetcaphorical concepts are central to culturalcogniition.Ethnographic studies an2d psychological experimentsinidicatethat coniceptualmetcaphors are expressed in numnerous forms of humanexpresand structureof metaphoremergsion, inicluding speech, ritual, narrative,and miaterialculture. Generalizationson the natutre ingfrom cognitive linguistic researchcan be used to develop methocdsfor reconstrulcting ancientnetcaphorsfromarchaeological evidence. In a preliminaitvcpplication, I argute thatpottery designsfromi2 the Mesa Verderegion of the AmericanSouthwestwere conceptualizedas textilefabrics, and suggest that conniectionsbetween these media der-ivedfron7 a worldviewgroundedin container imagery. The ability to decipher coniceptuial metaphors in prehistoric material culture opens up many n?ewavenuesfor including the role of worldviewin cultural evolution, an2d the discovery of cultural continuities between archaeologresear-ch, ical cultures and historic ethnolinguistic groups. Este artfculotrata de un2ificar recienites teor-fas sobre el significado culturalde la culturamaterial usanidola nocion de metifora conceptutal. Investigacionies en varias disciplinias sugieren que los coniceptos convencionalesson centrales en la cogmnetaf6ricos
nicion cultural. Estudios etnogrificos y experimentos psicologicos indican que las metaforas conceptuales estan expresadas en

variasformas de expresi6nhumana,incluyendolengucje, rituial, y culturamaterial.Las generalizacionessobre la natnarr-ativa, de la metifora que emergende la investigaci6nlingluisticacognitivapueden usarse para desarrollarmetouralezay estr-uictur-a dos para reconstruirmietdforas antiguas a partir de la evidencia arqueol6gica. En una aplicacion preliminar arguyo que los diseihoscerdmicosde la regi6nde Mesa Verde en el suroestenorteamnericanofuieron como textiles,y sugiero que conceptiualizados las conecciones entre estos miiedios La abilidad de decifrar den-ivani de utwaperspectiva anicladaen imagenqes de conitenedores. el descubrimmetdforasconiceptuales en.Ia culturamaterialprehist6ricaabte nuevas avenidaspara la investigaci6n,inicluyendo iento cle continutidades las cuiltlitrcs culturalesentr-e hist6ricos. arqueol6gicas y los grupos etniolinguisticos

his paper develops a methodologythrough whichsome of the cultural meaningsembedmaterialculturemight ded in archaeological be deciphered, andappliesit to the ancientPuebloan occupationof the Mesa Verderegion in the American Southwest.My approach is grounded in culTent understandings of mentalimageryin psychologyand linguistics, especially a cognitive phenomenon knownas conceptual metaphor. The traditional view sees it as a purelylinguisticembellishof metaphor mentthatis of littleconsequencefor theoriesof language and thought. In reality, conventional metaphorical conceptsare systematically expressed in everydaydiscourseandrevealthe fundamentally poetic natureof cultural cognition.In addition,conceptual metaphoris an image-based,nonlinguistic
T

phenomenon thatis expressedin materialcultureas well as language.In this paper,I arguethatarchaeologists can reconstruct ancient conceptual metaphors by examiningthe structure of figurative expressionsin archaeological materialculture,and by following a code-breaking proceduresimilarto that used in the deciphermentof ancient scripts. Based on the resultspresentedhere, I suggest that research on metaphor in material cultureholds great promisefor prehistoric archaeology. Theoretical Background The notionthatculturalcognitionis figurative,and in nature, has a long history especiallymetaphorical in anthropology. Frank Hamilton Cushing,one of the foundersof participant observation,wrote "I have

Scott G. Ortman * Crow CanyonArchaeologicalCenter,23390 County Road K, Cortez, CO 81321 AmericanAntiquity,65(4), 2000, pp. 613-645 Copyright( 2000 by the Society for AmericanArchaeology 613

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

614

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

found the Zuni arguesactualandessentialrelationship from similarityin the appearance, function,or other attributes of even genericallydiversethings" (1886:510-5 11). This conclusionwas also reached by Boas (1911:73), who commentedon "theuse of metaphorical terms in poetry,which, in rituals,are takenliterally, andaremadethebasisof certain rites." Levy-Bruhl(1926) arguedthatnon-Western modes of thinkingwere "prelogical" becausethey focused more on analogicalcorrespondences and relationthan on ships groundedin a "law of participation" the objective,logical consistencysupposedlycharacteristic of "civilized"rationality.Finally, LeviStrauss(1966), despite his critiqueof Levy-Bruhl, nevertheless agreedwith him on the pointthat"savage"thoughtis essentiallyanalogicalin character. havedismissedthese early Manyanthropologists the "savagemind"because attemptsto understand theycontrast analogicalnativethoughtwiththe supposedly logical, objective thought of Westerners. These views have also been criticizedas implyinga radicalcultural relativismthatis anathema to social science (e.g., Brown 1991). Surelyall humanshave brainsthatworkthe same way thatours'do, so the argument goes (for a historyof the "psychicunity" debate,see Shore 1996:Chapter 1). In recentyears havebegunto treada midcognitiveanthropologists dle pathbetween these two poles, arguingthatdifferences in culturalcognition do not derive from different mechanismsof thought,but fromdifferent of the world (Fernandez1986, conceptualizations and Levinson 1996; Keesing 1987; 1991;Gumperz Levinson 1996; Palmer1996; Salmon 1999; Shore arein factfigurative, 1996).Theseconceptualizations as earlyanthropologists noted,but areusuallytaken in a culture,includingour as literalby participants own. In short,figurative thoughtis basic anduniversal, butthe domainsof experienceused in figurative conceptualizations vary significantly,althoughnot withoutlimit or regularity, acrosscultures. Numerous ethnographicstudies illustratehow whichenableone thingto be conceptualmetaphors, in termsof another, are understood andexperienced expressedin variousformsof culturalbehavior(for a recent review, see Tilley 1999). Several of these studies (Bird-David 1990, 1992; Bourdieu 1973, 1990:271-283;Davidet al. 1988;MacKenzie1991; PrestonBlier 1987; Shore 1996:Chapter 11; Sillar 1996; Tilley 1999:Chapter 4; Turner1991; Walens arerevealedin multiple 1981) show thatmetaphors

modes of expression,from everydayspeech, to the structure of ritual,to events in sacrednarratives, to 1 Thisdivertheproduction, form,anduse of artifacts. sity indicatesthat metaphoris not merely a matter of poetic language,butrather is a matterof thought, whichprecedesandshapeslanguageas well as other forms of communication. In short, conventional metaphorical conceptsformthe foundationsof culturalunderstandings, and are createdand transmitted amongbeings withfundamentally poeticminds. ThePotentialof Metaphorin Archaeology Within archaeology,several theorists from seemingly disparateschools of thoughtare beginningto convergeon metaphoras an important concept for inferring the cultural meaningsencodedin artifacts. This trendfollows fromColinRenfrew'spromotion of "cognitive-processualarchaeology" (Renfrew 1994, 1998). One aspectof Renfrew'sprogram, the examination of "pre-modern" modes of thinking (Renfrew1994:5), has recentlygained momentum (see RenfrewandScarre1998) throughthe workof MerlinDonald(1991, 1998a),who arguesthatcognition has continuedto evolve in modernhumans, despiteminimalgenetic change,throughthe development of technologies for extending biological memory. Donald focuses on the emergence of speech andliteracyin his work,butrecognizesthat over the course of prehistoryhumanshave developed ever more complex means of storingcollective knowledge outside the brain using symbolic materialculture.Thus, artifacts,as expressionsof culturalknowledgeand as mnemonicdevices reinforcing it in the minds of their makers,gradually freedhumancognitionfromthe memorycapacities of individualbrains. issuefor Donald(1998b:187)arguesthata central cognitivearchaeologyis exactlyhow symbolicartifacts storememory,and exactly what it is thatthey "store." Onthebasisof ethnographic studies,he sugin the strict thanstoringinformation gests that,rather artifacts encodecultural logical formof a computer, knowledgein a fuzzy or analogicalway, following a "principleof perceptual and action-metaphor" (1998b:186). Donald's comments suggest that is a possiblesolutionto theproblem of how metaphor in cogachieve"external artifacts symbolicstorage" nitive archaeology. Metaphoris also migratingtowardthe centerof post-processualtheory.For example, in his recent

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHOR IN THE ARCHAEOLOGICAL RECORD

615

book Metaphorand Material Culture,Christopher Tilley (1999) providesa wide-rangingoverview of in the social sciences and on metaphor the literature is cencase thatmetaphor makesa strongtheoretical tralto the way meaningandmemoryareencodedin artifacts. To thisend,the most powerfulbody of eviby Tilley(1999:16-19)is thatemergdencereviewed psychology and linguistics. ing from experimental This evidence stands out for two reasons. First, in these fields areexplicitlyscientificin researchers outlook and are committedto conductingexperiments and formulatingtheoriesthat are consistent with empiricalresults (see especially Gibbs 1994; Lakoff1991, 1993;LakoffandJohnson1999). Secthese researchers ond, unlikeearlyanthropologists, have focused on the ways figurative thought is revealed in everyday English and have built on intuitionsin developing subtle and native-speaker designs.2The resultsof experimental sophisticated cognitiveanthrosuchstudieshaveconvincedseveral pologists (e.g., Palmer 1996; Salmon 1999; Shore is a centralcog1996;Sweetser1990)thatmetaphor nitiveprocess,andI believethisevidencemightconas well.Thatcognitive vincecognitivearchaeologists archaeology mightconvergeon andpost-processual metaphoras a common explanatoryconcept is a thought,especiallygiventheirradicallydifstriking ferentphilosophicalorientations. The Methodological Challenge. But even if it were accepted by all that "people do encode intothings" (Tilley1999:76), meanings metaphorical operationalizingmetaphortheory in archaeology would still pose a significantmethodologicalchalin material studiesof metaphor lenge. Ethnographic cultureareusuallygroundedin linguisticdatarelating to artifacts,and such data are not availableto It is one thingto claimthatmetaphor archaeologists. culture,butit is is centralto the meaningof material canreconto claimthatarchaeologists quiteanother structancient metaphorsfrom the archaeological recordalone. This problembecomes even moretrickywhen it in conuse metaphor is accepted thatanthropologists just as ceptualizingandwritingaboutothercultures their much as native subjectsdo in understanding own worlds.Functionalists conceptualizesociety as an "organism";interpretivists,culture as "text"; cultureas "syntax";and sociobiolostructuralists, as a collectionof "memes" (genes).The gists,culture underlie fact that metaphorical conceptualizations

theoriesdoes not necessarilymake anthropological them less valid or useful. In the end, we have no about complex choice but to think metaphorically But since both analystsand artisocial phenomena. the fundamental think metaphorically, fact-makers which metaphorsare questionis how to determine and which ones in the of the analyst, mind created 1999:36). minds (Tilley ancient characterized Both cognitive-processualand post-processual require us to answerthisquestion programs research previous attemptsto but unfortunately, somehow, from archaeological metaphors ancient reconstruct evidence have not been able to escape problems A notable posedby theabsenceof nativeinformants. example is Ian Hodder's(1990, 1992, 1994) argument that the chambertombs of Neolithic Europe by theirbuildersas "housesfor wereconceptualized the dead." In developing his argument, Hodder between (1990:142-156)listseightcorrespondences Neolithic houses and tombs, includingsimilarities and uses of in the dimensions,shapes,orientations, houses and tombs;the clusteringof both in "settlements,"and the common siting of tombs on top of make Hodearlierhouses. These correspondences seem plausible,but how could der's interpretation andothers we distinguish betweenhis interpretation we mightimagine,say,thatNeolithictomb-builders copied house-floorplans simply because they had models to work with. How no other architectural themwithoutaskingthebuilders mightwe establish, trulyreflect selves, whetherthese correspondences ancientmetaphorsas opposed to some more mundanephenomenon? I believe the best way to overcomethe nativetestimonyproblemis to considerpsychologicalandlinguistic researchon figurativethoughtmoreclosely, and focus on generalizations concerningthe strucThis tureof metaphor emergingfrom this research. is consideredbelow and will groundthe literature methodologyof the case study. What Is Metaphor, Really? The most systematicresearchto date on the nature in a has been conductedby researchers of metaphor relativelynew school of linguisticsknownas cognitive linguistics. The goal of this field is to develop a

theory of human language that is consistent with what is generallyknown aboutthe mind and brain fromdisciplinesotherthanlinguistics.Manyarchaewithgenhavesome acquaintance ologistsprobably

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

616

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

erative linguistics (e.g., Chomsky 1986; Pinker

possessa relatively 1994),whicharguesthathumans autonomous"languageorgan"and focuses on the in grammatiways linguisticsymbols are arranged concal sentences.Cognitive linguistics,in contrast, siders language to be inextricablybound up with broaderpsychologicalprocesses and focuses more each otherin the on how people actuallyunderstand context of natural discourse (Fauconnier 1997; Lakoff 1991; LakoffandJohnson 1999; Langacker anduse of 1991;Regier 1996).Since the acquisition languagederives from bodily experience,and this experience is filtered through nonlinguistic andmemimagery, processes,includingperception, reflect these ory, language should systematically As a result,cogbroader psychologicalmechanisms. nitivelinguistsstudylanguagebut gain insightinto the deeperworkingsof an embodiedmind. I suggest thatresearchfindingsin cognitive linguistics can help archaeologistsdevelop methods might through which ancientconceptualmetaphors evidence.Anthrobe deciphered fromarchaeological pologists, especially Boas (Aberle 1960), Sapir (1994), and Levi-Strauss(Leach 1970, 1976), have often looked to linguistics for theoreticalfoundations, but in recent years possible relationships between language and materialculturehave been the who havedocumented questioned by researchers in of "designgrammars" accountingfor inadequacy material-culture variation (Hardin1983;McCracken 1987). I suggestthatit is the adoptionof generative views of languagein these studies,andnot the connectionbetweenlanguageandmaterialculture,that aspect of language on is flawed. The appropriate whichtheoretical analogiesshouldbe builtis semantics, not syntax. Cognitive,linguistics focuses on semantics,andI believe a closer look at researchin thefoundation it needs thisfieldcangive archaeology for a more appropriate analogy between language and materialculture. Threebasic findingsof cognitive linguisticsare for this paper.First, both language and important in mentalimagery,or "mental thoughtaregrounded thatbegin as conceptualanalogsof representations vision, including experience," immediate, perceptual touch, taste, smell, hearing, and emotional states (Palmer1996:47). Second, the uniquehumanabilandprocessingmeantransferring, ity forproducing, sets of ing occurs throughmappings, or structured correspondences between domains of mental

imagery(Fauconnier1997). Third,althoughmany differentmappingmechanismsareknown,conceptual metaphoris the most common and important kind(Gibbs1994;Johnson1987;Lakoff1987, 1993; LakoffandJohnson1980, 1999; Sweetser 1990). To get an idea of what conceptualmetaphoris, in Engeverydayexpressions considerthenumerous ofjourneysto talkabout lishthatuse theterminology "he'sa life experiences("I need to move forward," lost soul,""we'refinally getting where we want to etc.). be," "our relationshipcrashed and burned," These are all surfaceexpressionsof an underlying, nonlinguistic metaphoricalconceptualizationthat idiom, canbe succinctlydescribed usingthefamiliar LIFE IS A JOURNEY (smallcapitalsareused to denote metaphoricalconcepts, after Lakoff and Johnson thesourcedomainis ajour1980).In this metaphor, ney, with a startingpoint, intendedgoals, obstacles to overcome, and a destination; and the target of this domain,onto which the conceptualstructure is more difis life, something that sourceis mapped, ficult to conceptualizedirectlyfrom bodily experience. In metaphor,the perceivablepropertiesand pointfor relationsof the targetdomaincorrespond, the source with of domain. People point, properties do not normallynotice that they are expressinga metaphorical thoughtwhenthey say thingslike "my "I'm stuckin a badmarcareer hascome a long way," a new headed in butEngdirection," or "I'm riage," the meaning immediately understand lish speakers the have internalized of suchstatements becausethey A use it conceptLIFE IS JOURNEY, and metaphorical to conceptualizelife experiencesand communicate aboutthem using metaphorical expressions.
GenercalProperties of Metcaphor

Carefulanalysis of large numbersof conventional metaphorical expressionshas revealedseveralgeneral propertiesof conceptual metaphor.Although everydayEnglishexpressionsprovidea convenient it is criticalto meansof illustrating these properties, languageis merelya surkeepin mindthatfigurative conface expressionof an underlying metaphorical ceptual system, and there is no formal difference betweenmetaphor as revealedby linguisticexpressions and metaphoras revealedby othermodes of humanaction, includingthe productionand use of material culture(Fauconnier 1;Lakoff 1997:Chapter 1993). The six propertiesreviewed in this section refer to metaphoras a cognitive process involving

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHOR IN THE ARCHAEOLOGICAL RECORD

617

mental imagery,which is nonlinguisticin nature. These propertiesare criticalbecause they can help determine how archaeological archaeologists expressions of conceptualmetaphorshouldbehave. Thefirstproperty is thatmetaphorDirectionality. ical mappingsare directional. The source domains areusuallygrounded in concretephysof metaphor ical experience,whereastargetdomainstend to be moreabstract. Metaphor enablesourpartialknowledge of a relativelyabstract phenomenon to be organized into a coherent image-schema or gestalt structure using a moreconcretesource,whichfacilitatesreasoningandcommunicating aboutthatphenomenon (Lakoff and Johnson 1980:56-68; 1999:Chapter 4). Inconceptual metaphor, thesemappings are asymmetrical, which means that abstract domainsarenot used to conceptualizedomainsthat are alreadyfairly concrete.For example,"springis still a long way off" and "themillenniumis finally behindus"expressthe conventional TIME metaphor IS SPACE, with physicalspace as the sourceandtime as the target.In this way of conceptualizingtime, futureeventsarein frontof a person,pasteventsare behind, physical distance correlates with the of time between two events, and so on "amount" (LakoffandJohnson1999:Chapter 10). In contrast, it is nonsensicalto say somethinglike "the ball is futureof me" (with the intendedmeaning"theball is in front of me") as there are no expressionsin everyday English or in any other language that expressSPACE IS TIME. Peopledo notdo thisbecause we alreadyhavea directbodilyexpelience of space, of space andit does not facilitateourunderstanding to conceptualizeit as time. TheSuperordinate Principle.This second propfrom the ertyfollows way humanbeings categorize perceptual inputinto kindsof experience.Much of our direct experientialknowledge is organizedin mental imagery as basic-level categories (Lakoff 1987;Rosch 1978;Varelaet al. 1991), whichrepresent the most inclusivelevel of abstraction thatcan be representedby a concrete image in the mind. Basic-level categories maximize the correlational structure of our direct interactionswith the world, and arethe firstcategoriesto be learnedand named by children(Rosch et al. 1976). More complex categorizationsthatorganizea mature person'sknowledge consist largely of subsets and supersets of basic-level categories.In English, chairs and birds are basic- level categoriesbecausethe averageper-

son canconjureup a mentalimageof a prototype for these concepts.Furnitureand animals, in contrast, are superordinate-level categolies because there is no single prototypefor these concepts that can be imagedin the mind(for an application of prototype theoryin materialculturestudies,see Miller 1985). It turnsout thatmetaphorical expressionsutilize basic-levelcategoriesthatare easy to image, while the actualmetaphorical conceptsthatgeneratethese variedexpressionsoccur at the superordinate level (Lakoff 1993). For example, in linguistic expressions of LIFE IS A JOURNEY, such as "myplanshave been derailed," the life is usuallydescribedas a car, train,boat, or plane attemptingto reach a destination, but the generalmappingfrom which all these specific expressionsderive involves the superordinate categoryvehicle. The Invariance Principle. This third property statesthatthe portionof the sourcethat is mapped onto the targetpreservesthe image-schematic strucit (Lakoff tureof the targetand does not contradict 1990, 1993). When aspectsof the sourceand target the targetoverridesand domaindo not correspond, limitsthoseaspectsof the sourcethatcanbe mapped onto it. For example,the metaphorTIME IS MONEY is commonlyused in our culture,as when we think about"budgeting," or"wast"saving "borrowing," ing" time (Lakoff and Johnson 1999:Chapter 10). However,notall aspectsof ourknowledgeof money are used in conceptualizing time. When we put thismoneyusuallyaccumoneyin a savingsaccount, mulatesinterest,suchthatthebalanceof the account grows over time, even if we do not add to the principal. But no one would arguethat if we set aside thatwe will miracusome time now for retirement thanwe directly lously havemoretime at retirement set aside for it. This aspect of our experience of money is not mappedonto time, because we experiencetime passingat a consistent,irreversible rate, and know that we cannot accumulatetime in the same way saved money accumulatesinterest.The invariance principle accounts for these kinds of restrictionson metaphorical mappings.Aspects of the source domain(principalaccumulatesinterest) that contradictthe structureof the target domain (timepasses at a consistent,irreversible rate)arenot mapped. The Constitutive Principle.The fourthproperty is thatconventionalmetaphorical conceptsareconstitutiveof thought(Gibbs 1994; Lakoff and John-

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

618

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

son 1980, 1999; Reddy 1993). Many social, psychological,andcosmologicaldomains,conceptualare taken as literally true by ized metaphorically, normal,unrein a culture,andstructure participants flective thinking.Forexample,Lakoff(1987:Chapand philosophers ter 11-15) has shownthatWestern analyticonceptualized scientistshave traditionally cal categoriesas containers.In this metaphor,the physicalentity as a bounded, of a container properties butalsoto reasonabout, areusednotonlyto describe, of entitiesin theworld.This is why theclassification containentitiesthatcanbe inside classicalcategories or outside of the category,but cannotbe both; and the classical syllogism (If X is in A andA is in B, then X is in B) is true because it follows from the of nestedcontainers. physicalproperties The world does not necessarilyhave this structure,but CATEGORIES ARE CONTAINERS iS so deeply inWestern culturethatit is oftenassumed embedded to representobjectivereality withoutintrospection way,the "outthere" in the world.To putthis another implied by ontological (what is) correspondences defineepistemic(whatyou metaphors conventional whichenablepeopleto canknow)correspondences, of reasonabouta targetdomainusing the structure themapped sourcedomain.Infact,it is nearlyimpossible to thinkaboutmost of ourbasic conceptswithof concepts outmetaphor. Ourliteralunderstandings causation, such as life, love, time, communication, the self, andthe cosmos areactuallyquiteimpoverhelpsus flesh outtheseconceptsso ished.Metaphor we can thinkaboutthemin moredetail(Lakoffand Johnson1999). of metaphor BlendedSources.The fifthproperty andTurner (1994) call derivesfromwhatFauconnier "middle spaces." In metaphor,the projection of onto the targetdomain is source domain structure mentalspace thatdefines motivatedby a "generic" betweenthe two domains pointsof correspondence at a higherlevel of abstraction (as in the superordinateprinciple).This genericspace governsthe projection of conceptual structure from the more concrete source onto the more abstracttarget.But and wrinklehere. Fauconnier thereis an important Turner(1994; Fauconnier 1997:Chapter6) have shownthatgenericspacescan alsobe set upbetween two domainswith relativelyequalconceptualstrucis not proture.In such cases, conceptualstructure fromthe betterknownsource jected asymmetrically to the lesser known target,because neitherdomain

is any more concrete than the other.Instead,coninto frombothdomainsis projected ceptualstructure space"thatcombinesdetailsof both a new "blended in an emergentstructure. An exampleof conceptualblendingcan be seen McGuirefinallybeatRoger "Mark in the statement, A literalreadMarisat the end of the 1998 season." impliesthatMarisandMcGuire ing of thisstatement were in a home-runcompetitionduring the 1998 baseballseason,which is impossiblebecauseMaris has been deadfor manyyears.However,since both seasons share McGuire'sand Maris'record-setting of a baseballseason, andboth the generic structure have relativelyequal inherentconceptualstructure, both seasonscan be projectedonto a blendedspace in which it becomes coherentto thinkof the two in competition. direct,contemporaneous Blendedspacesarecriticalfor two reasons.First, blended spaces can form new source domains for (Lakoffand projections morecomplexmetaphorical Johnson 1999:49). Second, blended spaces have thatis often physicallyimpossiemergentstructure ble, but conceptually coherent (Fauconnier and Turner1994). Throughblendedspaces, conceptual metaphor can become something more than a detailedmappingof real-world perceptualstructure Withblendedsourcedomains, fromsourceto target. that exists metaphorcan map conceptualstructure only in the mindandhas neverexisted in the objective, real world. Blended spaces are criticalfor the deciphermentof ancient metaphorsbecause they a meansof going beyondphysgive archaeologists and into the realm correspondences ical, real-world of purelyconceptualrelationships. The ExperientialPrinciple. The sixth and final is groundedin is thatconceptualmetaphor property bodily experience.An extensive body of research is deeply theconclusion thatmentalimagery supports of perception (Damasio1994; rootedin the circuitry Finke 1989; Kosslyn 1996; Tootellet al. 1988). As a result,metaphors only make sense when speakers knowledgeof the havesomeexperiential, perceptual sourceandtargetdomainsthatarelinkedmetaphorexpressions ically. In addition,novel metaphorical usually flow from the details of a person'sexperience of the sourceand targetdomains. Forexample,in recentyearsthe notionof a computer virushas emergedas a means of conceptualizing destructive computer programs written by Thepersonorpersonswho malevolent programmers.

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHOR IN THE ARCHAEOLOGICAL RECORD

619

.9 Wn; taX ,:.10 2 td

UT AZ
o
25
Kilometers

CO ~~~~~~~~~~~~
50

Figure 1. The Mesa Verde region. Numbered sites mark provenience of selected collections in Table 4. C)Crow Canyon Archaeological Center.

musthavehadfairlydetailed invented thismetaphor knowledge of both viruses and computers, since computerviruses "infect"othercomputersthrough in the the transfer of data,theyproduce"symptoms" affected machines, and computerscan be "disinagainstfuture"contaminafected"or "immunized" tion" (see Fauconnier1997:18-25). Furthermore, this metaphor wouldbe utterlynonsensicalto someone who hadno notionof whatvirusesandcomputers were and how they worked.Thus, directexperiential knowledge is an importantingredient in the of novel metaphorical inventionand understanding conceptsand expressions. The six properties of conceptual metaphor reviewedin this section are essentialto developing a methodologythat can determinewhethercorreculbetweendifferent classesof material spondences ture, or between artifacts and other phenomena experienced by ancient people, derived from they or fromsomethingelse. In particular, metaphor place significantrestrictionson how metaphorical in archaeological mateconceptsshouldbe expressed

rial culture.Let us now take a look at how these play out in a concreteexample. properties The Textile Metaphor in Mesa Verde Pottery Painting To illustratehow cognitive linguistic researchcan from ancientmetaphors decipher helparchaeologists record,I presenta case studythat the archaeological the mindsof thatcharacterized proposesa metaphor ancientPuebloansin the Four Cornersarea of the AmericanSouthwest.More specifically,I focus on the GreatPuebloPeriod(A.D. 1060-1280) occupation of the Mesa Verderegion, in southwest Colorado and southeast Utah (Figure 1; Lipe 1995; Varienet al. 1996;Wilshusenet al. 1997).3 many My proposalbuildson the correspondence havenoticedbetweenpaintedpottery archaeologists designs and woven textiles in this region.This correlationwas first noticed more than a centuryago. William Henry Holmes wrote in 1886 that "Ninetenths of archaic Pueblo, ceramic, ornamental designs are traceableto the textile art,and all show

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

620

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

the influence of textile convention"(1886:247). A few years later, Nordenskioldcommented that in potterydesigns "themotifsof the thirteenth-century first patternsexecuted in colours must be sought where of the textileindustry, amongtheproductions resultof a method is oftenthenecessary theornament of weavingorplaiting" (1990:86 [1893]). In thefirst Morriscommentedthatthe earhalf of this century, designsof thisarea"werestronglyinfluliestpottery enced by the antecedentand more familiarart of basketry" (1927:197); and Brew (1946:247) pottery,"manyof observedthat on eighth-century the designs were identicalwith those found on the decoratedbaskets."Awarenessof this relationship has continuedto the present(Brody 1991:61;Carlson 1982:208; Hays 1992:261-264; Hays-Gilpin 1977). 1995;Larralde The fact that textile details have been noted in Mesa Verderegion potterydesigns spanningmore thansevencenturiessuggeststhata long-term,conceptualrelationshipbetween these media is a reasonablepossibility.MesaVerderegiontextileswere usuallydecoratedthroughthe use of dyed weaving process.As a result,texelementsin theconstruction tiles were much more highly structuredby the mechanicsof variousweaving processes than pottery designs were by brushand paint.The weaving peoplelentthemknownto MesaVerde technologies selvesto geometric designs,buton a smooth,slipped, of designs andpolishedpotterysurface,anymanner been including representational could have painted, scenes, plantsand flowers, etc. Since such designs were not typicallypainted,andbothtextileandpottery designs were geometricin nature,it is reasonstylisticunity ableto proposethatthe centuries-long from projectionof the betweenthese mediaderived of textilesontopotterysurfaces. structure conceptual In otherwords,Mesa Verderegionpottersconceptualized these media as being equivalent and expressedthis equivalencein theirpotterydesigns. The evidence supporting this proposal will be adducedin the following pages. Analysisas Decipherment Metaphor The method I have used in exploring the textileto the metaphor hypothesisbearssome resemblance procedurephilologists have used in deciphering arequite ancientscripts. Althoughthetwo endeavors differentin their details, they share the important of not being groundedin the linear,deducproperty

havebeentrained tivelogic thatmanyarchaeologists to privilege (i.e., Hempel 1966). Nevertheless,the of ancientscriptsdoes producestable decipherment knowledgeof the past, and I suggest thatcomparably secureknowledgeof ancientmetaphorscan be by modelingoureffortson methodsof deciobtained pherment. A brief digressionon one recent decipherment will illustratemy point. LinearB, the scriptof the Mycenaean Greeks, was deciphered in 1952 by afterhe madethreecriticalassumpMichaelVentris tions aboutthe texts (for a full accountof this decipherment,see Chadwick1958). First,based on the in theinscripidentified characters of distinct number of LinearB stood tions, he proposedthatcharacters for pronounceablesyllables in whateverlanguage Second,a few LinearB characters theyrepresented. in the were physically similarto certaincharacters classicalCypriotscript,which hadbeen deciphered from bilingualinscriptionsmany years earlier,and Ventrisassumed that the phonetic values of these were similarin both scripts.Third,and characters thatthewords proposed Ventris mostcontroversially, werein anarchaic spelledoutby LinearB characters form of the Greeklanguage. These assumptionscould not be verified indeLinpendently,butVentrisneverthelessinterpreted ear B characters in accordance with these andfoundthatwhen this was done, the assumptions inscriptions beganto yield phrasesin a plausibleversion of archaicGreekthat also made sense in light of Linear of theircontexts.Thus, the decipherment B did not occur through the analysis of widely Ventris's observations. assumptheory-free accepted, whatthe datain LinearB tions actuallydetermined inscriptionswere;however,they did not determine wouldbehavecontheinterpreted characters whether of his model. This kind of with extensions sistently of (Hock is typical decipherment reasoning recursive and the applicaand Joseph 1996:94; Pope 1999), tion of this logic has producedstableknowledgeof the past in the form of readableancientscripts.So of characin decipherment, the way interpretations ters are obtainedis of no consequence;what does is whether whenapplied theseinterpretations, matter consistently plauto a corpusof inscriptions, produce if the interand can siblereadings, these onlyemerge pretationsof the charactersare basically correct (Chadwick1958:91). have been used in Similarlogic and procedures

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHOR IN THE ARCHAEOLOGICAL RECORD

621

Table 1: Archaeologicalmetaphoranalysis as a form of decipherment. Deciphermentof Ancient Scripts 1. Develop model of the language and kind of script (aphabetic, syllabic, logographic,or combination)representedin the inscriptions 2. Propose values of specific charactersbased on model, bilingual texts, and/orcontextualinformation charactersin inscriptions,using distri3. Examine interpreted bution, model, and contextualinformationto propose interpretationsof additionalcharacters 4. Determinewhetherresultantreadingsare consistent with general linguistic structure,the specific languageproposed, and contextualinformation ArchaeologicalMetaphorAnalysis 1. Develop model of the metaphoricalrelationshipexpressed in materialculture,specifying hypotheticalsource domain and targetmedium 2. Propose analogous featuresin targetmedium, based on image-schematiccorrespondencesbetween source domain and targetmedium 3. Examine distributionof analogous featureson targetmedium objects, and across time and space 4. Determinewhetherpatternsin analogous featuredata are consistent with general propertiesof metaphor,and with specific metaphorproposed

in Mesa evidenceforthetextilemetaphor marshaling Verdepotterypainting(Table1).The proposedrelationshipbetween these media explicitlyguided my designsthat pottery of MesaVerde search forfeatures to derivefromtextiles.The analogousfeaappeared turesI identifiedin the potterydesigns each exhibit withtextiles,butthereis correspondences numerous no way to prove, a priori, that they are indeed expressions.Nevertheless,I collected metaphorical dataon the ways these featurescombine in pottery of these featuresin designs, and on the distribution in patterns time and space, andidentifiednumerous the resultantdata that are consistentwith expectations derivingfrom generalpropertiesof metaphor Eventhoughthe proposed. andthespecificmetaphor usedanddatacollectedderivefromthe classification this model did not hypothesisunderconsideration, whether thesedatawouldbehavein accordetermine dance with extensionsof the model. Cognitivelinexpressions indicates thatbehavioral guisticresearch are highly structured, so much so that of metaphor the pottery-designdata are unlikely to behave in accordance withthis structure by chance.Therefore, I believe both the classificationand the hypothesis are supported by the results.
Collecting Evidence for the Textile Metaphor

The first step in adducingevidence for the textilethe history hypothesiswas to reconstruct metaphor of the proposedsourcedomain.The primarydecoratedtextilemediaknownfor theMesaVerderegion non-loom includecoiled basketry, plaitedbasketry, weaving,andloom-wovencottoncloth. Coiledbaskets were made by sewing successive circuitsof a

coil onto itself. Varicontinuous,outward-spiraling basketsweremade ousplantfiberswereused.Plaited by first plaitinga squaremat of yucca leaves, then forcingthe mat througha circularhoop and fastening the ends of the plaiting stripsto it. Non-loom weavings utilized warpsand wefts in creatingnarrow articlesincludingbelts, tumpbandsor carrying straps,aprons,and sashes. Non-loom articleswere made of Apocynumsp. fiber, dog or human hair, of yucca cordage,cotton,andvariouscombinations these. Finally,cottoncloth was woven on uprightor backstrap looms using heddles thatcould pick sets Thesewerealso warp-weft of warpssimultaneously. weaves and were madeexclusively of cottonyarn. even in the arid Southwest, the Unfortunately, is not recordof these four industries archaeological as completeas one wouldlike. Althoughtextilecolcliff dwellings are lections from thirteenth-century collections extensiveand well-studied,comparable contextsarequite fromeleventh-andtwelfth-century rare. However,ancient Puebloantextile industries throughto havedevelopedfairlyconsistently appear out the greater San Juan river drainage (Kent 1983a:124-125), of which the Mesa Verderegion part,so the comproencompassesonly the northern all significant texmise chosenwas to firstinventory tile mediaanddecorativetechniquesin Mesa Verde regioncollections,andthento look beyondtheMesa to Chaco Canyon Verderegion itself, in particular (Dutton 1938; Judd 1954; Pepper 1996), Chinle Wash(AdovasioandGunn1986;Magers1986),and the PrayerRock District (Morris 1980), for additionalexamples.Whenthislargerareais considered, enough materialis availablefor a historicalrecon-

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

622

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

4a

C)a)O

4-4~
0 ~ ~~ ~~ ~~~~~~~~~~~~~~~~~~~~~~0 U~~~~~~~~~~~~~~~~~~~~~~c

00

00 a) n Cm 0 uCO 0CC

Sa)

G,0

0~~~~0 &~

u~~~~~~~~~~~~
0 0 0
a)

~ ~ ~ ~ u000 *Qa)~~~~~~~~~~~~~~~~-~~~~4)

~~~~~~~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~0 ON~~~~~~~~~~~~0 ~ ~ ~~ ~ ~~~~~~~~~o~~~~~~S.S

a)

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHOR INTHE ARCHAEOLOGICAL RECORD

623

struction of these industries.A summary of this into textilesthe way theywerepaintedon pottery,so the methodof weaving designs was not transferred is given in Table2. reconstruction Next, I examinedcomplete potteryvessels and to the sequenceused in paintingthem. anduse of analogous Toexaminethedevelopment drawings,and descriptions publishedphotographs, in pottery designsacrosstimeandspace,colfea- features of textilesin developinga list of pottery-design tures that:a) occur in specific decorativezones of lectionsof bowl-rimsherdsfromtrashdepositshave withtreeMostof theseareassociated to haveanalogues beenanalyzed. servingbowls;b) appear decorated so the assemblagescould be in specific weaving techniques;and c) are usually ring dated structures, of the ceramicdata.Table4 lists identifiableon sherds.The 25 analogousfeatures I datedindependent metaphor- theanalyzed alongwiththeirlatestassoassemblages, to conventional arecomparable identified phaseto which date,thechronological ical expressionsthatwouldbe compiledin studying ciatedtree-ring of rimsherdsanalyzed, figurativespeech.These featuresare listed in Table eachis assigned,the number cal3 accordingto the sourcetextile industryand deco- andthe numberof sherdsincludedin frequency datafrommorethan4,000 sherds Attribute rativezone of a potterybowl on which each occurs. culations. sites, totalingmore archaeological expres- from29 different The featuresarelabeledusing metaphorical are includedin observations, sions that specify the textile detailsthatare argued than 100,000 separate The sites are locatedwithinthe Sand to have been expressedin potterydesigns. Most of this database. in this paper.Detailed Canyon locality, the Ute Mountainpiedmont,and these featuresare illustrated and Colorado, features, andanassess- McElmoDrainageareasof southwest of theanalogous descriptions Utah(Figure1). In genmentof theirdateof inventionin textiles,is given in in severalareasof southeast an appendix available on the SAA website eral,I have followed recenttrendsin the interpretadata,which see multipleclustersof saa.org/publications/amAntiq/AQAbstracts/tion of tree-ring (www. dates as evidenceof recyclingratherthanremodelAq65-4/appendix). Eachanalogousfeaturewas definedon the basis ing (Bradley 1993; Varien 1999:Chapters5-6). between painteddecorationon Breaksbetweenphasesarebasedon the distribution of correspondences datesfromthese sites. a specific decorativezone of a potterybowl andthe of latesttree-ring kindof texThe analysisfocusedon rim sherdsbecausesuch of a particular structure image-schematic in tab- sherds are more numerousand more consistently are enumerated tile. These correspondences thanwhole vessels, andtheypreserveeviular form in the appendix.To illustratewhat these recovered Figure2 dence of the decorationappliedto all four decoraanalogousfeaturesarearguedto represent, vessel (exterior, rim,interior of a relativelysimplemap- tivezonesof theoriginal presentsaninterpretation Also, the depositionof sherds ping, Feature16. Figure 3 illustratesthe source of margin,andinterior). thanthe placFeature16, andan exampleof its expressionin pot- seems to havebeen moreunreflective properties ing of certainvessels in burials,or the decision of tery.In Figure2, formalandinteractional rim, exterior,containobjects,etc.) shared ancient people to leave vessels inside abandoned (interior, andWilshusen (Lightfoot1993;Schlanger by plaitedbasketsand potterybowls are linked via structures collectionsfrom midandthislinkageenables 1993). As a result,rim-sherd thegenericspacecontainers, samplesof the dens shouldprovidemorerepresentative the transferof additionalconceptualstructure, correspondences themselves, from the mental the potteryused duringan occupationthan whole of sherdsis that sur- vessels. The primarydisadvantage imageryof a plaitedbasketonto a pottery-bowl face. However,notice that some aspects of plaited there is greaterpotential for misidentificationof thatwould be clear on a completevessel. presumablybecause attributes basketryare not transferred, a three-value such a transfer would violate the invariance princi- In orderto minimizethiskindof error, was used in ple. Forexample,potterydesigns were addedto the system (present,absent,indeterminate) surfaceof a completed and dried vessel and were recordingeach analogousfeature,and sherdsfrom paintedby subdividingthe field into smallerareas the same vessel were identifiedto providea means Veryfew examplesof conandthenfillingthemin (Rohn1977:163-164;Shep- of checkingconsistency. ard 1963:296). Plaited basket designs, in contrast, tradictory,determinateobservations were found were built up line by line, as plaiting strips were amongthe conjoiningsets in the database,suggestwere rare. identifications addedto the growingmat.Designs were not woven ing thatinaccurate

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

624

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

.o

o
X4->
x

b-

Ct~~~~~~~~~~~~~~~~~~~~~~

<~~~~~~~~~~~~~~~~~~~~~~~~~~~E :
U4'IC =

4
=

o) U

U o/

U;

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHOR IN THE ARCHAEOLOGICAL RECORD

625

Generic: Containers

Source: Plaited basketry *containobjects


-exterior :

Target Pottery bowl contain objects


exterior

*rirn *interior *coloredplaiting strips * *twill-plaiting *centereddesign *squaremat draped over osier ring *design line by line

~~~~~~~~~~~~~~~~~rim
nterlor i solid lines counterchangedpositive and negative space centered, all-over design rectangularto circular layout *painted *design by outline and fill

Figure 2. Diagram of mapping represented by analogous feature 16. Properties of plaited baskets and pottery bowls that are common to the generic concept containers define the mapping, which enables the transfer of additional conceptual structure from plaited basketry to pottery surfaces. Note that elements of plaited basketry that do not correspond to pottery are not projected.

MetaphorHypothesis Extensionsof the Textile reviewedearlier The generalpropeltiesof metaphor make predictionsabout the structureMesa Verde designsshouldhaveif theywereindeedmatepottery Datacollected rialexpressionsof a textilemetaphor. hypothunderthe guidanceof the textile-metaphor

esis exhibit numerouspatternsthat are consistent arediscussed These patterns with these predictions. to whicheach to theproperty in thissectionaccording relates. pattern Principle.TheidentifiedanalTheSuperordinate ogous featureslisted in Table3 relateto fourdifferplaitedbasketry, ent textilemedia-coiled basketry,

Figure 3. Colored, twill-plaited basket and analogous pottery design illustrating feature 16, colored twill-plaited design. a) basket from Mesa Verde, after Morris & Burgh 1941:Figure 36b, C)Crow Canyon Archaeological Center; b) pottery from Alkali Ridge, Utah, after Brew 1946:Figure 71b.

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

626

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

00 00 0

00s

00

00

NC C C NC~~~~~~~~~~~~~~~~~~~~~~ CO cO cO~~~~~~~~~~~INCNCN > 6 =-~~~~5-3 = 6 == = = =

00 ,

00 00

00 00

NC

> > > c~~~~~~~~~~~~~~~

D oo It) o > Cs <


t -<

aC)

00

>

>

>

00

00S

) r-

W)

CO)

CD ) * cO

C) Ist ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~O (nc O l 1> ?t

cr

nnCn nC C1 C1 C1

tNIr It; >)

>

> oo

>oW

cea)Ntnm

\C

CA oo

'>nNt b

?t

COL, CO
03
cn;

ct~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~c

a)

-cj

CfCf

00

n 00

00?

0 00 nn 0

00o0

o~~~~X= -o~~~g3X

EX

vvvvvv

>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Ev 0 ?

oXXoooooo

<S,
? m t m m m m m m m

CO
t

E~~~~~~~~~~~~~~~~~~~~~~~~ 0, = =

0 Ca

CO

~ ? ~~~~~~~~~~~~~~~~~~~~~~~ <a C 001


--m

at)

( .O
rm c -CT

Ev Ev F
S

<(

<

n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i .if-)

SS S

vE

SS
C

vE
C

HE

oc~ooa vE vE
S

SS

vE

>>>>>>-t < o ioo S

vE

Clm m

S C Cl

~o
=0S
o

HE

o
S

oo~
S
m

vHE

S
m

o ooa vE vE
S SS
m m C 'IC]

' _0, O_\ X~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~W O\ _I\ t O\ O\ O\ 0 0YYY

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

RECORD METAPHOR INTHEARCHAEOLOGICAL CONCEPTUAL ? ? Xnon-loom \C

627

00

o o
X0
$--4 $--4

o=X
= =
s.,

3
O

444
0 o 000

gX
X

t
g

fabrics, and loom-woven cotton cloth. These arebasic-levelcategories,becausea concrete image of a typicalexampleof each kindof weaving in contrast,is a is easily createdin the mind.Textile, superordinate category,because there is no single concretementalimage thatcan be generated to rep-

$--4

-14

u u u u

in resentall the textileindustries thatparticipated


mappingswith pottery.Thus, the generalmapping cateappearsto have been from the superordinate expresbutspecificmaterial gory textilesto pottery, sions of themappingareatthebasiclevel, consistent
with expectations of the superordinate principle.

TheDirectionality Twopatterns supConstraint. portthe notionthatthe relationship betweentextiles and potterywas asymmetrical; thatis, thatconcepOt-

- oo CA e,\ X__

CIA0 CIA a,\ cn

It r_ 00 00 00 00

rt~00000000 V<000000\~0000

v
X

< ;> rH m b2

> ;>;> ;>


Iv >Nt

;
X

X
H >D

pottery,and not the otherway around.First, many of thecorrespondences features thatdefineanalogous 0 \ 0 could not have been transferred from r-. r-. o potteryto textiles.Themorethan100correspondences thatdefine the analogousfeaturesin Table 3 derive from five differentaspectsof textile design. Each correspon00 00 mn 00 00 00 dence is labeled (a-e) in the appendix table according to the aspectof textiledesign to which it relates, andthe numberof correspondences relatingto each aspectis tabulated in Table5. o;> ;> o \b ? \Although certainmotifsanddesignlayoutscould o m NNNc m > have N been ? ? invented firstin potteryandsubsequently tt transferred to textiles,manyof the correspondences in Table5 areveryunlikelyto havebeen summarized inventedin pottery.Category(d) representscorren0 ? ? b?n o

>tual

occurred from textiles onto projection primarily

.0

spondences that map incidental details imparted by

F
-14

CQ
,
4

,,, 3 0 <> 0 XX o different For example, when a weaving processes. i


triangle is sewn into a coiled basket wall, a leg that is diagonal to the rim and travelsagainstthe work
m t
t

v m

Xm

Z zmv

4O m

z;

m W =

directionwill necessarilyappearjagged due to its travellingin opposition to the stitch-slant(Figure 4a); butwhen a triangleis executedin twill-tapestry weave on a loom-woven fabric, it is a vertical leg
Table 5. Summaryof Correspondences. Number Exhibited
17

0404;. ;,, u 0~

u
4Z4

~ ~

Type of Correspondence
(a) Use of colored weaving elements
>

.
=
m

<

.
X

=3(b) Colored motifs ^ t (c) Design structureand layout


(d) Incidental details of weaving processes

ar
oo o oo ;; oo oo
Z E_

r-

~ ~

(e) Surfacetextures
listed in appendix

8 59 11 14
109

X ,

E< FE E EH= ?? ??- ?Total correspondences co co co co co co co

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

628

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

e
Figure 4. Coiled basketry and twill-tapestry cotton cloth triangles, and ticked triangles on pottery. a) basket from Canyon del Muerto, after Morris & Burgh 1941:Figure 41q, C Crow Canyon Archaeological Center; b) design on cloth from Mesa Verde, after Kent 1983a:Figure 8.12 (also illustrated in figure 12a); c) pottery design from the Cowboy Wash Area (Collection 19), triangles illustrate feature 12, coiled/non-loom triangle; line at top is on the rim, illustrating feature 1, colored rim coil; true background in band illustrates feature 10, coiled color design; d) pottery band design from Badger House, Mesa Verde, overall design illustrates feature 18, pre-cotton non-loom band design, after Hayes and Lancaster 1975:Figure 126; triangles illustrate feature 12, coiled/non-loom triangle, C Crow Canyon Archaeological Center; e) pottery band design from the La Plata District, triangles illustrate feature 25, twill-tapestry triangle; overall design illustrates feature 22, twill-tapestry band design, after Morris 1939:Figure 67.118.

on a potterysurface.As examples,Features5 and 7 lines, which runaroundthe vessel use thin-framing spacesbetween, withwiderunpainted circumference to representthe darkintersticesbetween coils in a coiled basketwall (Figure5); Feature15 uses broad lines" to representthe relationship "hatchure-filled these details were not intentionally woven into tri- between individualstripsof yucca and the overall angles in differenttextile media but were unavoid- interval-shift pattern in twill-plaited ring-basket able consequencesof differentweavingprocesses. designs (Figure6); andFeature21 uses background allareasof centered, whichfills unpainted (e) includescorrespondences hatchure, Inaddition, category the twill-ribsof a twilled inpaint over designs, to represent textures rendered thatrepresent textile-surface (parallelto the warpson an uprightloom) thatwill (Figure4b). jagged, due to the twill-rhythm appear These subtledifferencesin triangleswoven in various mediawereactuallyexpressedin potterydesigns as tickedtriangles,with the tickingon the appropri13 and25; Figure4c-e). Notice that ateleg (Features

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHORIN THE ARCHAEOLOGICAL RECORD

629

Figure 5. Coiled basketry wall and analogous pottery framing pattern illustrating features 5-8, coil interstices, alternating colored and uncolored coils, colored coils and interstices, and stitch-marks. a) basket from Horse Rock Ruin, SE Utah (Adovasio & Andrews 1990, Type III, Container 2), ? Crow Canyon Archaeological Center; b) pottery from Long House, Mesa Verde, after Cattanach 1980:Figure 132b.1, C) Crow Canyon Archaeological Center.

Figure 6. Plain twill-plaited basket and analogous pottery design illustrating feature 15, uncolored twill-plaited design. a) basket from Mesa Verde, after Nordenskiold 1990:Plate XLV.2 [1893], C) Crow Canyon Archaeological Center; b) pottery from Shiprock, New Mexico, after Lister & Lister 1978:Figure 19, C) Crow Canyon Archaeological Center.

cottonfabric(Figure7). Thatincidentaldetailsand surface textures were also mapped onto pottery designs arguesstronglythat the directionof mapIt is very difficult pingswas fromtextilesto pottery. new weavingtechto conceiveof weaversinventing existingdetailsof painted to translate niquesin order or incidendesignsintonew surfacetextures pottery taldetailsof motifson textiles.

the directionality The second patternsupporting is thatnone of the 25 analogousfeatures constraint to beingattested in pottery designswasinvented prior each is arguedto derivefrom. in the textile industry of the 25 Table6 presentsfrequencies(percentages) analogousfeaturesfor sherd collections dating to eight chronological phases of the Great Pueblo Period.4 Certaincells of the tablehave been shaded to indicatephasesin which the sourceof each analogous featureis attestedin the textile record (see Appendix and Table 1). This table shows that, althoughthere are a few features(especially 2, 23 to haveexistedfor some time in and24) thatappear to pottery,none of textilespriorto being transferred

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

630

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

Figure 8. Example of "non-grammatical" continuation of an interior design derived from cotton cloth onto a pottery bowl rim, C Crow Canyon Archaeological Center.
i t/

Figure 7. Twill-tapestry cotton fabric and analogous pottery design illustrating feature 21, twill-rib background. a) cloth from Grand Gulch, after Kent 1957:Figure 97a; b) pottery from the La Plata District, after Morris 1939:Plate 318a, C Crow Canyon Archaeological Center.

the analogousfeaturesappearsin more than trace amountsduringphases for which the textile counIf we allow of these featuresareunattested. terparts for a few misclassifications and intrusivesherds,it becomes very likely thatnone of these featureswas expressedin potterydesigns priorto being invented in the relevant textilemedium.Also, in almostevery case, the frequenciesof analogousfeaturesincrease in the phases thatfollow this initial transfer. These inherthe notionthat,dueto the greater datasupport the entstructure of textiledesign,textilesrepresented

sourcedomainandpotterythe targetdomainof the metaphor. TheInvariancePrinciple.This principleis supported by the fact that, for features derived from coiled basketry,rims are mappedonto rims, walls Only two of onto walls, andinteriorsonto interiors. themorethan4,000 analyzedsherdsexhibitrimdecfrom designderived thatcontinueaninterior orations wovencottonfabricsontotherimsurface(Figure8). prinThese isolatedcases do violate the invariance ciple,becausetheedges of woventextilesdo notcorrespondto the rim of a potterybowl, but they are exceedinglyrare(we will see laterhow the mapping of warp-weft weaves onto pottery interiors was accomplished). Also, despite the fact that some rim decoration elements,suchas ticksanddashes,werealsopainted in framingspaces, thereis not a single example of X's andzig-zagsbeingusedin this way.Why should be used in framing some elementsof rimdecoration andnot others?This apparently mysterious patterns makesperfectsense in light of the invarirestriction ance principle.Ticks and dashes could be painted between framinglines as well as on rims because colby wrapping bothweremadein coiled basketry the coil as it was sewn around oredstitchingmaterial onto the growing basket. If the colored stitching materialwas insertedon the final coil, it formed a rim tick; if it was insertedfartherdown the vessel the wall, it becamea tick markon a coil. In contrast, technique,fromwhichX andzig-zag false-braiding rim decorationsare arguedto derive,could only be appliedto the rim of a coiled basketand could not be done in the basketwall. Thus,it wouldhave vio-

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHOR IN THE ARCHAEOLOGICAL RECORD

631

Table6. Analogous Featuresin Mesa VerdePotteryDesigns, A.D. 1060-1280. Phase (years A.D.) 4 3 114011101180 1140

Source Industry of Feature Zone wherefeature occurs Analogous feature

1 10601090

2 10901110

5 11801230

6 12301250

8 7 1250- 12601260 1280

Coiled Basketry Rims 42.3 (1) Colored rim coil 0.0 (2) Rim stitching 0.0 (3) Rim stitch gaps rim 0.0 (4) False-braided Interior vessel walls/margins 3.9 (5) Coil interstices (6) Alternatingcolored & uncoloredcoils 0.0 0.0 (7) Colored coils and interstices 0.0 (8) Stitch-marks Interiordesigns 26.2 (9) Coiled color designs 2.0 (10) Coiled texturedesigns 3.8 (11) Coiled terrace 0.0 (12) Coiled/non-loomtriangle Exteriordesigns 6.8 (13) Coiled surface texture Plaited Basketry Interiordesigns (14) Simple plaited designs (15) Uncolored twill-plaiteddesigns (16) Colored twill-plaiteddesigns Exteriordesigns (17) Exteriorselvage band designs Non-loom Warp-Weft Weaves InteriorDesigns (18) Pre-cottonnon-loom band designs (19) Post-cottonnon-loom band designs terrace (20) Plain-tapestry Loom-woven Cotton Cloth Interiordesigns (21) Twill-ribbackground band designs (22) Twill-tapestry (23) All-over twill-tapestrydesigns terTace (24) Twill-tapestry triangle (25) Twill-tapestry Non-textile Designs (26) Non-basketryrim decorations (27) Non-basketryframingpatterns (28) Mixed weave & non-textile interiordesigns Number of local painted vessels examined

27.9 6.0 0.0 0.0 1.1 0.0 0.5 0.5 14.2 0.6 1.0 1.9 1.8

19.4 10.5 0.0 0.0 0.0 0.4 1.1 0.4 14.0 0.4 0.0 2.5 3.6

6.0 32.0 2.0 3.3 10.3 9.5 17.2 4.8 18.6 18.6 1.4 0.0 1.6

2.9 46.5 10.3 4.2 3.7 13.0 21.5 5.4 9.1 6.3 0.0 0.0 0.4

2.2 44.0 8.2 3.8 6.0 3.6 38.1 2.0 7.2 7.1 0.0 0.8 0.6

1.8 48.2 11.7 5.1 5.1 10.2 30.8 6.5 9.5 12.6 1.0 1.4 0.8

0.1 42.3 8.2 8.0 6.1 7.0 34.2 4.2 5.5 9.0 0.4 1.1 0.3

5.1 35.9 10.3 0.0

7.3 27.6 16.3 0.0

13.8 25.6 28.6 0.0

2.1 11.6 10.5 2.3

3.8 1.5 10.5 2.0

0.5 5.3 14.5 6.5

3.7 1.5 5.9 9.7

2.1 1.1 8.6 11.1

15.4 0.0 3.1

16.3 2.4 4.1

4.9 1.0 0.6

4.2 4.2 6.0

0.8 8.3 12.4

1.0 8.7 11.5

0.7 5.9 12.7

1.4 5.8 14.2

0.0 0.0 0.0 0.0 0.0

2.4 0.0 0.0 0.0 0.0

3.0 0.5 0.0 0.0 0.0

8.4 6.3 0.0 0.0 1.8

18.0 22.6 2.3 5.3 4.0

16.4 21.7 4.8 1.6 4.0

12.6 27.4 2.2 4.1 2.7

18.1 26.5 2.8 7.7 3.3

0.0 1.8 2.4

0.0 0.0 9.0

0.0 0.0 6.5

0.0 1.9 4.9

0.4 2.3 2.6

0.9 1.5 4.9

0.0 1.8 0.7

0.7 2.1 4.3

59

220

334

187

266

364

259

1671

Note: Numbersare percents. Shadedcells indicate presence of featurein textile record.

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

632

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

c
Figure 9. Coiled basket compared with "grammatical" and "non-grammatical" framing patterns in pottery. a) basket from Long House, Mesa Verde, after Cattanach 1980:Figure 420, ? Crow Canyon Archaeological Center; b) "grammatical" pottery rim sherd from Sand Canyon Pueblo, rim illustrates feature 4, false-braided rim; framing pattern illustrates feature 7, colored coils and interstices, and feature 8, stitch-marks, ? Crow Canyon Archaeological Center; c) unattested, "non-grammatical" pottery rim sherd due to placement of feature 4 in framing spaces, ? Crow Canyon Archaeological Center.

latedthe invariance principleto mapa false-braided rim coil onto framingspaces (Figure9). The Constitutive Principle.The primarypattern the notion thatpotterydesigns were litsupporting erallyconceptualized as textilefabricsis thatpottery designs unrelatedto known weaving processes, or thatcombineweavingprocessesin impossibleways, of the neverconstitutemorethana smallpercentage totalcorpusof designsin the analyzedassemblages.

This pattern holds for rim decorations, framingpatterns,andinterior designs(Table6). Thesedataindicate that"non-textile" designsdid not become more popularover time. Howevei; the fact that some do It shows thatartists consistentlyoccuris important. did occasionally imagine designs that were poor applicationsof the metaphor, or were simply unrelated to it, but that such designs seem to have been evaluatednegativelyby otherpotters,and were not often emulated. The textile metaphor therefore appearsto have placed limitationson the invention and evaluationof new potterydesigns. Blended Sources. Features 18 through 25 are argued to derive from warp-weft weaves, woven eitherwithor withouta loom (Table3). It is straightforward to imaginehow detailsof coiled andplaited basketrymight be mappedonto pottery-vesselsurforwarpfaces,buthow wouldthisbe accomplished weft weaves, which generallydo not havebasketor potteryshapes?The answeris conceptualblending. It canbe shownthatwarp-weft weavesweremapped onto pottery-vesselsurfacesby first being blended andthatentailwitheithercoiled orplaitedbasketry, mentsof theseblendingswere carried overintopottery designs. Non-loom warp-weft weaves, such as belts, aprons,sashes, andtumpbands,were long, narrow, and flexible, with the warpaxis following the long travThedesignon sucharticles dimension. naturally eled alongthe longerwarpaxis. Inuse, theseobjects were often formedinto bandsthatwrappedaround the waist or head. By conceptually mapping the warpsof these bandsonto the coils of a coiled basnon-loomarticlescould ket,banddesignson narrow, be blendedinto the interiorof a coiled basket,and transferred to potthis blendedimage subsequently tery (Figure10). Figure 11 illustratesthis mapping,which shows that points of correspondencebetween non-loom apronsandcoiledbaskets,andadditional conceptual structure uniqueto each medium-such as theplain tapestry weavingmethodandtherimfinishingtechontoa new, niqueof coiledbaskets-were projected to pottery. Inthis blendedsourcedomainfor transfer case, coiled-basketwall surfacetextureand a plain tapestry-weave banddesign were mappedonto the This blendedsourceis same pottery-vesselinterior. impossibleon a single woven object, but it is conceptuallycoherentin termsof mentalimageryand its manipulation.

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHORIN THE ARCHAEOLOGICAL RECORD

633

MffP

~M

Figure 10. Non-loom apron, coiled basket, and analogous "blended" pottery design. a) apron from northern Arizona, after E. B. Sayles photo in Kent 1983b:Figure 42, C Crow Canyon Archaeological Center; b) coiled basket from Long House, Mesa Verde, after Cattanach 1980:Figure 420, C Crow Canyon Archaeological Center; c) vessel from Sand Canyon Pueblo, after Lee Schmidlap drawing, rim illustrates feature 4, false-braided rim; framing pattern illustrates feature 7, colored coils and interstices; motif illustrates feature 20, plain-tapestry terrace; band design illustrates feature 19, post-cotton nonloom band design; C Crow Canyon Archaeological Center.

of cottonandheddletechnology Theintroduction thecreative increased aroundA.D.1100dramatically resourcesof pottersfor innovativemappings.Wide fabrics,in which the weft axis was not significantly shorterthan the warp axis, began to be woven on weredeveloped looms, andnew weavingtechniques and surthatenablednew motifs, design structures, face textures to be created. How were these new ideas transferredto pottery? Band designs were non-loom wovenintocottoncloth,butunlikenarrow

articles,the bandin cotton cloth traveledalong the to the warps.So to map a weft axis, perpendicular on cotton cloth (Feain twill-tapestry band woven fabric wouldneedto be conthe ture22) ontopottery, the wefts mapped into a tube, and ceptually"rolled" An example of a basket. of a coiled onto the coils clothandin potin cotton band design twill-tapestry showshow in 12. This example Figure teryis shown on cottoncloth a banddesignwovenin twill-tapestry and a coiled basket surfacetexturecould be transferredonto the same pottery-vesselinterior. Centered,all-overdesigns were also woven into cotton cloth, and it appearsthat such designs were mappedonto potterythroughthe conduitof plaited a square mat basketry, (Figure13).Inplaited basketry was wovenfirst,andthenit wasforceddownthrough, andfastenedonto, a circularhoop. By mappingthe cotton cloth onto warp and weft of a quadrilateral the plaitingstripsof thismat,the cottonfabriccould for the plaitedmat and thus mapped be substituted onto pottery to create a centered, all-over design derivedfrom cotton cloth. Figure 14 illustratesthis kind of mapping,and again shows how correspondencesbetweenloom-wovencottonclothandplaited uniqueto basketry,as well as conceptualstructure a to create physically each medium,were blended impossible but conceptually coherent source for interior. mappingonto a pottery-bowl here just a few represent The examplesdescribed blendings of the morecommonandstraightforward thatMesa Verdepottersset up in imaginingpottery designs. By considering entailments of the two via blending processes discussed here, "banding" (see Figures5, 10, 12) and"draping" coiledbasketry via plaitedbasketry(see Figures3, 6, 7, 13), we can adduceevidencethattheseblendingsreallyoccurred Framing patterns in themindsof MesaVerdepotters. that frame-band designs are arguedto derive from are surfaces,and in this framework coiled-basketry of non-loomartiin mappings coherent conceptually cles andweft-axisloom-wovenbanddesignsin cotton cloth onto pottery.In contrast,framingpatterns in mappingsof plaitedbasketry wouldbe incoherent and centered cotton cloth designs onto pottery, cannotbe cresurfacetextures becausecoiled-basket atedin plaitedbasketry. So, accordingto this model, combining framing patternsand draped, all-over design layouts on potterywould violate the invariance principle. of complex theoccurrence Table7 cross-tabulates

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

634

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

,,,,,,.....
___________ _____ _

Generic: Textiles *vegetal *woven *fabric Source2:Coiled basketry design line by line * stitching B foundation/coils coils form bands
stitching material

A ,*container

Source,: Non-loom apron *design line by line * *weft


*warp

*ends can connect


*dyed weft
e 0

'dyed

*plaintapestry

BlendSource3

colored coils

treglarsymetr ~~ -plain~~~

~
en \B

~~~~~~~~ zi-aeo
dSou*irn

i *pa intedsiil etr byout baddesignbyotieadfl

Figure 11. Diagram of mapping represented by pottery vessel in Figure 10. Points of correspondence in the generic space are not duplicated in the two source domains for the sake of brevity, and a second generic space (containers) governing the mapping of the blended source onto pottery (as in Figure 3) is left out for the sake of clarity. Non-loom fabrics and coiled basketry are related via the generic space textiles. Points of correspondence between non-loom fabrics and coiled basketry are projected, along with additional conceptual structure unique to each medium, onto a blended space, which serves as the source domain for metaphorical projection onto pottery. Note that elements of the blended spaceof at do not correspond to pottery are not projected.

design and"draped" with"banded" framing patterns A.D. 1140,whencomlayoutsforsherdsdatingafter 5The table weremostcommon. patterns plexframing shows that, indeed, complex framingpatternsare rarelycombinedwith drapeddesign layouts in the analyzedcollections.A few examples of complex framingand drapedlayouts do occur,but these are no more prevalentthan the non-textileand mixedweave interiordesigns in Table 6, suggesting that were selected againstby potters such combinations Thereis no physicalreaandneverbecamepopular. son why complex framing patterns and all-over designs should not have been combinedin pottery designs. On average,framingpatternsonly take up the top two centimetersof the interiordesign field, leavingplentyof roomfor all-overdesignsto be creYet,thiscomatedovertherestof thevessel surface. binationdid not normallyoccur,and the reasonfor

this remainstotallyobscureuntilit is consideredin light of blendedsources. is Principle.This finalproperty TheExperiential that by two patterns.First,it is apparent supported Mea Verdepotterswere not stronglyinfluencedby designs on importedvessels, becausedemonstrably nonlocalsherdsarequiterarein GreatPuebloPeriod Mesa Verde sites and decrease in frequencyover time (Lipe 1995:158;Ortman2000). If therewere evidence of significantculturalinfluencefrom outside the region, it would raise the possibility that potterydesignsinvented MesaVerdepottersadopted elsewheresimplybecausethedonorculturewasperceived as prestigious.If this were the case, Mesa Verde potters could have emulated designs that derived from a textile metaphor in some other Puebloanculture,but mightnot have possessed the themselves.But since MesaVerdepotters metaphor

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHOR IN THE ARCHAEOLOGICAL RECORD

635

__

Figure 12. Twill-tapestry cotton cloth, coiled basket, and analogous "blended" pottery design. a) cloth from Mesa Verde, after Nordenskiold 1990:Plate L [1893], ?) Crow Canyon Archaeological Center; b) basket from Mesa Verde, after Morris & Burgh 1941:Figure 27j, 18b, ?) Crow Canyon Archaeological Center; c) vessel from Sand Canyon Pueblo, after Lee Schmidlap drawing, rim illustrates feature 2, rim-stitching, and feature 3, rim stitch gaps; motif illustrates feature 25, twill-tapestry triangle; and band illustrates feature 22, twill-tapestry band design, ?) Crow Canyon Archaeological Center.

vessels do not appearto have experiencedimported veryoften, this possibilityseems unlikely. Second, intra-regionalspatial patternssuggest textileindusthattheexposureof pottersto different tries influencedtheir creativity.Potterycollections datingto theA.D. 1250-1280 periodhavebeen analyzed from 11 differentlocalities within the Mesa Verderegion (locationsgiven in Figure 1). Analogous featurefrequencieswere calculatedfor these 11 localities, using the same proceduresdiscussed for Table 6 (see Note 4). Since sample sizes were small for severalof the localities,these rawproporusing error forrandom-sampling tionswereadjusted empiricalBayesian methodsoutlinedin Robertson (1999). This proceduremodified observed analoin lightof the samplesfrom gous featureproportions which they were calculated,the global distribution of proportions for each analogousfeatureacrossthe of analeleven localities,andthe local distributions in adjacent localities.6 Posproportions ogousfeature for the five teriorBayesianestimatesof proportions loom-woven cotton cloth features, and six coiled

basketryfeaturesthatreachedtheirpeakpopularity duringthe mid-thirteenthcentury,were then analyzed by principalcomponentsanalysis(PCA). Loadings for the 11 analogousfeatureson the first principal component are given in ascending orderin Table8. The firstprincipalcomponentrepresentsthe dominantpatternof variationin analofromthe 11localities.These proportions gousfeature loadingsshowthatanalogousfeaturesderivingfrom loom-woven cotton cloth tend to have a positive score, while featuresderivingfrom coiled basketry Figure 15 is a thematicmap tendto load negatively. thatlabels the 11 localities accordingto theirscore on the first principalcomponent.This map clearly gradientin potterydesigns illustratesan East-West features in the east thatemphasizescoiled basketry cottonclothfea(negativescores), and loom-woven is also turesin thewest(positivescores).Thispattern maps of individualanaloapparentin distribution andin PCAsof therawfreproportions, gous-feature quency data that are not presentedhere. However, analysisof Bayesianestimatesprothe multivariate

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

636

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

Figure 13. Twill-tapestry cotton cloth, plaited basket, and analogous "blended" pottery design. a) cloth from Grand Gulch, after Kent 1983b:Plate 10, (C Crow Canyon Archaeological Center; b) basket from Mesa Verde, after Morris & Burgh 1941:Figure 37a, (? Crow Canyon Archaeological Center; c) pottery bowl from Long House, Mesa Verde, after Cattanach 1980:Figure 163g, interior illustrates feature 23, all-over twill-tapestry design, ? Crow Canyon Archaeological Center; d) pottery bowl from Mug House, Mesa Verde, after Rohn 1971:Figure 173e, exterior illustrates feature 17, exterior selvage band design, ? Crow Canyon Archaeological Center.

of vides theclearestandmostconciserepresentation trendsin the existing data. Based on the principlethatculturetraitstend to fromtheirplacesof origin(Sapir be diffuseoutward 1949:410-420 [1916]), this analysis suggests that analogous features derived from coiled basketry tendedto originatein Colorado,and those deriving fromloom-wovencottonclothtendedto be invented in Utah. Direct evidence of cotton cultivationand weaving,includingcottonseeds, loom anchors,cotton threadand weaving tools, is also much more common in southeastUtah, in both cliff-dwellings andopen-airsites (LindsayandDykman1978;Lipe et al. 1988;Walker 1960;Lipeet al. 1960;Thompson 1977). In contrast,comparableevidence does not occurin anyof theSandCanyonLocalityorMcElmo Even in cliff Drainagesites in the potterydatabase. dwellings on Mesa Verdeproper,the evidence sug-

gests thatcotton was not cultivatedlocally (Brown 1975; Rohn 1971). However,the raw materialsfor basketmakingcan be found everywhere,andbone awls suitablefor use in basketweaving, fragments sherdsare and basket-impressed of coiled basketry, theregion.Thus,potoftenfoundin sitesthroughout ters everywherewere well acquaintedwith coiled basketry. These data suggest that potters in the western MesaVerderegionhad moredirectexperiencewith cottonweavingon looms, andfocused on this experience in conceptualizingnew pottery designs. In contrast,pottersin the easternarea had less exposure to cotton weaving, leading to a creativefocus on coiled basketry.This correspondencebetween local emphases in textile industriesand derivative in different areas pottelydesignssuggeststhatpotters theyknewwell in imagdrewupontextileindustries

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHOR IN THE ARCHAEOLOGICAL RECORD

637

Generic: Textiles *vegetal *woven *fabric


_____ ____ ____ ____

* A Source2:Plaited basketry 'design line by line xplaiting strips - 'twill-plaiting x isquare plaited mat .dyed plaiting strips drapedover osier ring *~~~~~~~~mat attachedto ring 0 exterior selvage band Target: Pottery bowl regular geometry radiating,all-over design solid motifs backgroundhatchure geometric center, circular periphery no framing pattern
\ ticked rim

--

*container

Source: Cotton Cloth *design line by line ~ * *warp& weft * *twill-tapestry cloth *rectangular cloth1~ -rectangular *dyedwefit
*twilt-ribs -twill-ribs

-.

Blend Soure Source3 Blnd, -selvages

*- 0

exterior band design *painted *design by outline and fill

Figure 14. Diagram of mapping represented by pottery vessels in Figure 13. Points of correspondence in the generic space are not duplicated in the two source domains for the sake of brevity, and a second generic space (containers) governing the mapping of the blended source onto pottery (as in Figure 3) is left out for simplicity. Cotton cloth and plaited basketry are related via the generic space textiles. Points of correspondence between cotton cloth and plaited basketry are projected, along with additional conceptual structure unique to each medium, onto a blended space, which serves as the source domain for metaphorical projection onto pottery. Note that elements of the blended space that do not correspond to pottery are not projected.

extensionsof the textilewithnumerous ining potterydesigns, consistentwith the experien- accordance metaphorhypothesis,and patternsin the resultant tial principle. used in definingthese the assumptions datasupport Metaphor? Whythe Textile believethatPOTTERY analogousfeatures.I therefore The data presentedin the case study raise the fol- IS A TEXTILE describesan ancientmentalphenomeof the anal- non thatreally was sharedamong Mesa Verdepotlowing question:whatis the probability databehavingin accordancewith all ters and that is decipherablefrom archaeological ogous-feature of metaphor, andwithentail- remainsalone, withoutthe benefitof nativeconsulthesegeneral properties specifically,purelyby tants. mentsof the textilemetaphor is whyMesa Whatthecase studydoes notaddress chance? Even if it cannot be proved deductively, evidence thatsupports Verde potters should have conceptualizedpottery thereis variedand abundant in MesaVerde pot- designsas textilesurfaces.Thistopicdeservesmuch theexistenceof a textilemetaphor tery painting, and I believe this evidence is too more attentionthancan be given to it here, but for detailed to havearisenby chance,orby wishfulthink- now it should be sufficientto say that potteryand to havebeen used as sourcedomains in this textilesappear reported ing on my part.Manyof the patterns and were not for additional metaphorsin Mesa Verde culture. study were completelyunanticipated even consideredas expectationswhen I began col- Thus,stylisticunitybetweenpotteryandtextilesmay lecting dataon Mesa Verdepotterydesigns. Never- have been but one expressionof a more complex theless, the analogous feature data do behave in worldviewgroundedin the imageryof containers.

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

638

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

Table7. InteriorDesign Layout vs. Multiple-LineFramingPatterns. Complex Framing Pattern Absent Present Total Banded Layout 372 (47%) 414 (53%) 786 (100%) Draped Layout 518 (92.5%) 42 (7.5%) 560 (100%)

Total 890 456 1346

The form anddecoration of the MesaVerdekiva is a salientexample.Althoughthey appearto have been componentsof residentialarchitecture (Cater andChenault 1988;Lipe 1989;Ortman 1998;Varien 1999),MesaVerdekivasfollowedrigidarchitectural conventions and contained numerous features of The typicalkiva was symbolicor ritualimportance. circularand subterranean, was enteredthroughthe roof, and containeda small circularhole or sipapu the "earth thatrepresents navel"or emergenceplace in all modemPueblocultures.In addition,the walls of MesaVerdekivaswere occasionallydecorated in with the decoration of pottery ways thatcorrespond bowls (Martin1936; Morris 1991). Even framing patternswere occasionallypaintedon the pilasters or roof supportsof kivas (e.g., Morris1991:Figure 5.24). Finally,the roofs of Mesa Verdekivas were of concentriccircuitsof timbers,which constructed formeda hemispherical vegetal surfacethatis perceptuallysimilarto a coiled basket. This combinationof a "coiledbasket"roof with bowl" walls in the kiva suggests that tex"pottery tiles andpotterywerelinkedin additional metaphorical conceptsthatdefinedthe Mesa VerdePuebloan often incorworld.Houses and religiousstructures

poratecosmologicalmetaphors into theirform and decoration (Parker PearsonandRichards1994),and in certainmodem Pueblo culturesthe kiva is consideredan expressionof a cosmos comprisedof an earth-bowl below and an overturnedsky-basket above(BrodyandSwentzell 1996;Swentzell 1990). Thisconceptualization mapsverywell ontotheform anddecoration of MesaVerdekivas(Figure16) and may be evidence for continuity in worldview betweenMesaVerdePuebloansandcertainmodem groups (Ortman1999). If indeed the Mesa Verde kivawas a metaphorical expressionof anearth-bowl andsky-basket for cosmos, one possibleexplanation the textilemetaphor in MesaVerdepotterypainting would be thatpotteryand textiles were considered to be complementary partsof a largerwhole. This exampleshouldbe sufficientto illustrate how much moremightbe learnedabouttheconceptualsystems of Mesa VerdePuebloansfrom additionalstudy of metaphorical expressionsin theirmaterialculture. Discussion and Future Directions Before discussingimplicationsof this study,a few disclaimers arenecessary. First,I do notmeanto sugalone can accountfor gest thatthe textile metaphor the overall development of Mesa Verde pottery of interaction netdesigns. Obviously,the structure the socialconworks,the workof influential potters, ditions within which potters worked, and other artalso psychologicalaspectsof interaction through the case study roles.Nevertheless, playedsignificant theclaimthata particular strongly supports metaphorof potterydesigns exerteda ical conceptualization strongand systematicinfluenceon the evolutionof MesaVerdepotterypainting, even if thisinfluenceis only one dimensionof a morecomplicatedstory.

Table 8. VariableLoadings, First PrincipalComponent. Analogous Feature (2) Rim stitching (8) Stitch-marks rim (4) False-braided (10) Coiled texturedesigns (21) Twill-ribbackground (7) Colored coils and interstitces (25) Twill-tapestrytriangle (3) Rim stitch gaps (24) Twill-tapestry terrace (22) Twill-tapestryband designs (23) All-over twill-tapestry designs *Percentof total varianceexplained = .460 Source Industry Coiled Basketry Coiled Basketry Coiled Basketry Coiled Basketry Loom-woven Cotton cloth Coiled Basketry Loom-woven Cotton cloth Coiled Basketry Loom-woven Cotton cloth Loom-woven Cotton cloth Loom-woven Cotton cloth First PC Score* -0.969 -0.836 -0.822 -0.709 -0.596 -0.579 -0.519 -0.439 0.519 0.589 0.686

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHOR INTHE ARCHAEOLOGICAL RECORD

639

First PC Score
Bayesian Estimates 0.72 to 1.41 OOto 0.)

0.72

-0.68 to 002 1.38to -068

UT

co

25 Kilometers

50

Figure 15. Spatial variation in Mesa Verde pottery designs, A.D. 1250-1280. Based on the variable loadings in Table 8, darker shading indicates higher frequencies of loom-woven cotton cloth features, and lighter shading indicates higher frequencies of coiled basketry features. C Crow Canyon Archaeological Center.

Second,I also wantto makeclearthatin the texit was the mental imageryof textile tile metaphor, design thatwas mappedonto potterysurfaces.The use of blended sources, and the paintingof textile surfacetextureson pottery,shows that it was not of texphysicaltextiles,buttheconceptualstructure in as imagery existed only mental tile media,which mapped was potters, that minds Mesa Verde of the onto pottery surfaces.Metaphorin the decorative artsis morethancopying acrossmedia;it is the use of a source medium to of the conceptualstructure structured of a less inherently thedecoration structure targetmedium.In the case of Mesa Verdepottery emerdesigns,theresultsof thisprocesscanproduce that is not reflected in actualtextile gent structure specimens,but is neverthelesscoherentin termsof andmanipthewaystextileimagery wastransformed ulatedin the mindsof potters. Withthesepointsin mind,I believethecase study supportsthe theoreticalnotion that cognitive linbuild midguistic researchcan help archaeologists realm theideational dle-range theoryfordeciphering record(sensu Cowgill 1993). in the archaeological

the Archaeologistswill neverbe able to reconstruct totality of ancient belief systems, because ancient beliefs need not have been expressedin behaviors traces.Butevenif thetotalthatleavearchaeological ity of ancientconceptualsystems cannotbe reconstructed from archaeological remains alone, the ability to identify some conceptual metaphors archaeologically opens up many new areas of Two mostof whichhaveyet to be explored. research, of these will be brieflydiscussedhere in closing, in researchon concepfurther the hope of stimulating record. in the archaeological tual metaphor First, the ability to infer aspects of conceptual systems archaeologicallyallows us to examine the role of ideology and worldviewin the evolutionof culture.I suspectthatwhenthis is done,we material will findthatthisprocessis muchmorecomplicated theoriessuggest.Forexample, thanmanyprevailing (Neiman 1995; Teltser archaeologists evolutionary style can that 1995)haveargued evolutionin artifact theory of be explainedusing an adaptationof the in biology characters evolutionary selectivelyneutral 1970).Inthismodel,it is assumed (CrowandKimura

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

640

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

0Zi

Figure 16. The Mesa Verde kiva as an expression of the earth-bowl and sky-basket cosmos (after Morris 1991:Figure 5.24, Swentzell 1990:Figure 3-1). C Crow Canyon Archaeological Center.

that stylistic variantsare selectively neutral,from which it follows that innovationoccurs effectively at random,andthe transmission of variants to other artifactmakersderivesprimarily from the structure in the population. of interaction The archaeological case discussed in this paper suggests that the assumptionof selectively neutral is not always true.The historyof stylistic variation analogousfeaturesin Mesa Verdepotterydesigns illustrates thatinnovationwas anythingbut random in this style tradition and that"non-textile" designs were systematicallyselected againstvia the textile metaphor. To put this anotherway, the conceptual systemsof MesaVerdepottersexerteda strongand systematic influenceon the evolutionof MesaVerde pottery designs. So even if stylistic variantshave negligiblefitnessbenefitson a biologicallevel, certain variationsdo resonatemore stronglywith preand in cases like existing culturalunderstandings, Mesa Verdepotterypaintingthe resultingselective canbe verystrong. pressures Thecase studythussug-

gests thatconsideration of cultural cognitionin addition to interaction and biological fitness might significantly improve evolutionary theory in archaeology. Second, the ability to infer ancient metaphors from archaeologicalremainsopens up whole new areasof researchon continuitiesin conceptualsystems between archaeologicalculturesand modem ethnolinguistic groups.If conceptualmetaphor is a common denominator underlying language and material research thatcompares ancientconculture, ceptual systems, as revealed by archaeological remains, with historic conceptual systems, as revealedby languagesof relateddescendant groups, shouldshed considerable light on culture-historical relationships betweenthetwo (Ortman 1999).Inthis age of NAGPRA,methodsfor establishingcultural NativeAmericanbeliefs affiliationthatincorporate andscientificperspectives wouldbe simultaneously value.In addition,the identification of considerable of cultural continuitieswouldenablearchaeologists

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHOR IN THE ARCHAEOLOGICAL RECORD

641

Phoenix. Bird-David,N. AnotherPerspectiveon the 1990 The Giving Environment: CurrentAnthropolEconomicSystemof Hunter-Gatherers. ogy 31:189-196. 1992 Beyond "TheOriginalAffluentSociety":A Culturalist 33:25-47. Anthropology Current Reformulation. Boas, F. Indian In Handbookof NorthAmnerican 1911 Introduction. Languages,editedby F. Boas, pp.5-83. Bulletin40, Bureau WashingInstitution, of AmericanEthnology.Smithsonian ton. Art. Dover,New York. 1955 Primitive Blinman,E. 1988 AdditiveTechnologiesGroupFinalReport.In Dolores Report,compiled ArchaeologicalProgram,Final Synthetic C. K. Robinson,and G. T. Gross, pp. by D. A. Breternitz, of the Interior, Bureauof Recla53-102. U.S. Department mation,Denver. while Bloomer,W. W. of thisresearch wereconducted Portions Acknowledgments: supportedby a National Science Foundationgraduateresearch 1989 Moon House:A PuebloIIIPeriodCliff Dwelling Complex in Southeastern Utah. Unpublished M.A. Thesis, fellowship. Additionalfunding was providedby the Colorado StateUniversity, Washington of Anthropology, Department HistoricalSociety, the Utah Division of State History,the BalPullman. of Anthropology atArizona lantineFamilyFund,the Department Center. Adele Bourdieu,P. andCrowCanyonArchaeological StateUniversity, 1973 The BerberHouse. In RulesandMeaning,editedby M. Bigler, Brian Brownholtz,CarolynCurrie,KristinFangmeier, Douglas,pp. 98-110. Penguin,New York. Sue RobinLyle,JamieMerewether, Roberts, BonnieHildebrand, 1990 TheLogic of Practice. StanfordUniversityPress, Palo Dylan Schwindt,VernaSmith,andMaggieThursall contributed Alto. countless hours towardcompilationof the ceramics database. Bradley,B. A. the line drawingsof textilesandpottery at a PaulErmigiotti rendered 1993 Planning,Growth,and FunctionalDifferentiation JourPueblo:A Case StudyfromSW Colorado. Prehistoric critical insights relatedto this researchhave vessels. Important nal of FieldArchaeology20:23-42. De Boer, Betsy Brandt, been providedby Eric Blinman,WaiTen Anne MarshallChristner,George Cowgill, Donna Glowacki, Brew,J. 0. 1946 The Archaeologyof Alkali Ridge, SoutheasternUtah. Bill Lipe,AndreaParkes, JimPotMichelleHegmon,IanHodder, Papersof the Peabody Museum of Archaeologyand EthWashMark Dorothy Varien, Salmon, Enrique ter,ColinRenfrew, University,Cambridge. nology, No. 24. Harvard andRich Wilshusen.I would like to thank Brody,J. J. burn,LaurieWebster, Crow Canyon Soil Systems, Inc., The Ute MountainUte Tribe, 1991 Anasaziand Pueblo Painting.Universityof New MexStateUniversity, Edge of the Washington Archaeological Center, ico Press,Albuquerque. CedarsStatePark,and the Anasazi HeritageCenterfor permis- Brody,J. J., andR. Swentzell 1996 ToTouchthe Past: ThePaintedPottetyof the Mimbres sion to analyzeceramiccollections.Finally,I would like to espeR. WisemanArt Museum,Minneapolis. People. Frederick at cially thankMarkVarienandMelitaRomasco,my supervisors of this research, even Brown,D. E. CrowCanyon,for theirunflaggingsupport New York. McGraw-Hill, 1991 HumanUniver-sals. when thereseemed to be no end in sight. Brown,J. 1975 TechnologicalAnalysis of PrehistoricCotton Textiles References Cited Confrom MancosCanyonandAssociatedEnvironmental siderations. In The 1974 Johnson-Lion Canyon Project, Aberle,D. F. Reportof InvestigationI, assembledby Paul R. Nickens, 1960 The Influenceof Linguisticon EarlyCultureand Per91-140. Mesa Verde Research Center, Department of inHonor sonalityTheory.InEssaysin theScienceof Culture Universityof Colorado,Boulder. Anthropology, editedby G. E. Dole and R. L. Cameiro, of LeslieA. Whiite, Carlson,R. L. pp. 1-29. ThomasY. CrowellCo., New York. Ceram1982 The PolychromeComplexes.In Southwestern Adovasio,J. M., and R. L. Andrews pp. Review,editedby A. H. Schroeder, ics, A Comparative An Cache: 1990 TheHorseRockRuin(42SA10550)Basketry 201-234. The ArizonaArchaeologist15.ArizonaArchaeoTechnolView of RegionalPuebloPerishable Unparalleled logical Society, Phoenix. atthe55thAnnualMeetingof the Socipresented ogy. Paper Cater,J. D., and M. L. Chenault ety forAmericanArchaeology,Las Vegas. Lore54(3):19-32. Southwestern 1988 KivaUse Reinterpreted. Adovasio,J. M., andJ. Gunn G. S., Jr. InArchaleolog- Cattantach, 1986 TheAntelopeHouse BasketryIndustry. 1980 Long House, Mesa VerdeNational Park, Colorado. atAntelopeHouse,Don P.Morris,Senior ical Iinvestigations ResearchSeries7-H. NationalParkService, Archaeological SerDirector, pp. 306-397. NationalPark andProject Author D.C. Washington, D. C. vice, Washington, Chadwick,J. Billman,B. (editor) 1958 TheDeciphermnent of LinearB. CambridgeUniversity of the Ute MountainPied1998 The Puebloan Occutpation

to to bringthe fullness of linguisticreconstructions the archaeologicalrecord. their aid in interpreting historicallinguisticsandarchaeology Traditionally, but minimallyinterhave providedcomplementary acting views of the past, where linguisticevidence has been strongon culturebutweak on locatinghistoricalevents in time and space; whereasarchaeological evidencehas providedpreciselocationsand information. datingforevents,butmuchless cultural Metaphortheory should enable new kinds of collaborationbetween linguists and archaeologists, which would probablymake significantcontribudimensions of thecultural tionsto ourunderstanding of humanprehistory.

inArchaeology22, 12 vols. Soil SystemsPublications mnont,

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

642

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

Press,Cambridge. Chomsky,N. 1986 Knowledgeof Language:Its Nature,Origin and Use. Praeger, New York. Churchill, M. J. 1996 LetterReporton 1996 Excavationsat Woods Canyon Colorado. Report Pueblo(5MT11842),Montezuma County, on file, CrowCanyonArchaeological Center,Cortez. Cowgill, G. L. 1993 DistinguishedLecturein Archaeology:Beyond Criticizing New Archaeology. American Anthropologist 95:551-573. Crow,J. F., andM. Kimura 1970 An Introduction to PopulationGeneticsTheory.Harper Row,New York. Cushing,F. H. of Zuni Cul1886 A Study of Pueblo Potteryas Illustrative tureGrowth.4th AnnualReport,Bureauof AmericanEthWashington, nology,pp. 467-521. Smithsonian Institution, D.C. Damasio,A. R. 1994 Descartes' Error: Emotion, Reason, and the Human New York. Brain.Grosset/Putnam, and K. Gavua David,N., J. Sterner, 1988 Why Pots are Decorated. Current Anthropology 29:365-389. Donald,M. University 1991 The Origins of the Modem Mind. Harvard MA. Press,Cambridge, 1998a Hominid Enculturation and Cognitive Evolution. In Cognitionand MaterialCulture:TheArchaeologyof Symbolic Storage,editedby C. RenfrewandC. Scarre, pp.7-18. McDonaldInstituteMonographs, Cambridge. Thoughts. 1998b Material Culture andCognition: Concluding In Cognition of SymandMaterialCulture: TheArchaeology bolic Storage, edited by C. Renfrew and C. Scarre, pp. Cambridge. 181-187. McDonaldInstituteMonographs, Dutton,B. L. 1938 Leyit Kin, A Small House Ruin, Chaco Canyon,New Mexico.Monograph Series No. 7, Vol. 1. School of American Research, Santa Fe, and University of New Mexico Press,Albuquerque. G. Fauconnier, Uni1997 Mappingsin Thoughtand Language.Cambridge versityPress,Cambridge. Fauconnier, G., andM. Turner Report9401, 1994 ConceptualProjectionandMiddleSpaces. of CognitiveScience, La Jolla. UCSD Department J. W. Fernandez, The Play of Tropesin 1986 Persuasionsand Performances: IndianaUniversityPress,Bloomington. Culture. 1991 Introduction: Confluents of Inquiry. In Beyond TheTheory inAnthropology, editedby Metaphor: of Tropes J. W. Fernandez, pp. 1-16. StanfordUniversityPress,Stanford. Fewkes, J. W. 1909 Antiquitiesof Mesa Verde National Park,Spruce Tree House. Bureauof AmericanEthnologyBulletin41. SmithD.C. sonianInstitution, Washington, 1911 Antiquitiesof Mesa Verde National Park,CliffPalace. Bureauof AmericanEthnology Bulletin 51. Smithsonian D.C. Institution, Washington, Finke,R. 1989 Principles of Mental Imagery.The MIT Press, Cambridge,MA. Gibbs,R. W., Jr.

1994 The Poetics of Mind: FigurativeThought,Language, and Understanding.CambridgeUniversity Press, Cambridge. J. J., and S. C. Levinson Gumperz, 1996 Introduction:Linguistic Relativity Reconsidered. In RethinkingLinguisticRelativity,edited by J. J. Gumperz UniversityPress, and S. C. Levinson,pp. 1-20. Cambridge Cambridge. B. G., andC. D. Breternitz Harrill, 1976 Chronologyand CulturalActivity in JohnsonCanyon fromTree-Ring Data.JourCliff Dwellings:Interpretations nal of FieldArchaeology3:375-390 Hardin,M. A. 1983 ApplyingLinguisticModels to the DecorativeArts:A Consideration of the Limits of Analogy.SemiPreliminary otica 46:309-322. Hays, K.A. a Theoryof 1992 AnasaziCeramicsas TextandTool:Toward CeramicDesign "Messaging." UnpublishedPh.D. disserof Anthropology, Universityof Arizona, tation,Department Tucson. Hays-Gilpin,K. A. of thePuebloan Region:New Views 1995 ArtandArchaeology from the Basement.Museum Anthropology19(3):47-57. Hempel,CarlG. 1966 Philosophy of Natural Science. PrenticeHall, Englewood Cliffs, CA. Hock, H. H., andB. D. Joseph 1996 Language History,Language Change, and Language to Historicaland ComparaAn Introductiotn Relationship: Berlin. tive Linguistics.Walterde Gruyter, Hodder,I. of Europe.Basil Blackwell,Oxford. 1990 TheDomestication 1992 Burials, Houses, Women and Men in the European Neolithic. In Theoryand Practice in Archaeology, edited by Ian Hoddler,pp. 45-80. Routledge,London. and Meaning:The Example of Neolithic 1994 Architecture Houses andTombs.InArchitecture and Order: Approaches Pearson to SocialSpace,editedby M. Parker andC. Richards, pp. 73-86. Routledge,London. Holmes,W. H. Art in Its Relationto the Develop1886 A Studyof the Textile ment of Form and Ornament.Bureau of Ethnology, 4th WashAnnualReport, Institution, pp. 189-252. Smithsonian ington,D.C. Johnson,M. 1987 The Body in the Mind: The Bodily Basis of Meaning, Imagination, and Reason. University of Chicago Press, Chicago. Judd,N. M. 1954 The Material Cultureof Pueblo Bonito. Smithsonian MiscellaneousCollections,Vol. 124. Washington, D.C. Keesing,R. M. 1987 Models, "Folk"and "Cultural": Paradigms Regained? Models in Languageand Thought, editedby D. In Cultural HollandandN. Quinn,pp. 369-394. Cambridge University Press,Cambridge. Kent,K. P. 1957 The Cultivationand Weavingof Cottonin the Prehisof theAmertoricSouthwestern UnitedStates.Transactions ican PhilosophicalSociety 47:457-732. Philadelphia. of Traditional Shifts in the Structure South1983a Temporal westernTextileDesign. In Structure and Cognitionin Art, edited by DorothyK. Washburn, pp. 113-137. Cambridge UniversityPress,Cambridge. 1983b PrehistoricTextiles of the Southwest.School of Amer-

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHOR IN THE ARCHAEOLOGICAL RECORD

643

of the CrowCanyonArchaeologicalCenter 1, Cortez. ican ResearchPress,SantaFe. SanJuan: Conditions of theNorthern 1995 The Depopulation Kosslyn,S. Archae1200s.JournalofAAnthropological in the Turbulent Debate. 1996 ImageandBrain:TheResolution of theImagery ology 14:143-169. The MIT Press,Cambridge. D. S. Dibble, and K. M. Anderson Lipe, W. D., F. W. Sharrock, Kuckelman, K. A. 1960 1959 Excavations, Glen CanyonArea. University of Resultsof TestingandReportof 1997Exca1997 Preliminary Papers,Vol. 49. Salt Lake City. UtahAnthropological vations at Yellow JacketPueblo (Site 5MT5), Montezuma Lister,R. H., and F C. Lister County, Colorado [HTML title], http://www.crowof New MexicoPress,Albu1978 AnasaziPotteiy.University canyon.org/ResearchReports/YJP/Report.htm. querque. Langacker, R. W. The CognitiveBasis of MacKenzie,M. 1991 Concept,Image, and Synmbol: Objects:StringBags and Genderin Cen1991 Androgynous Berlin. Grammar. Moutonde Gruyter, tral New Guinea.HarwoodAcademicPress,Melbourne. Lakoff,G. Fire,and DangerousThings:WhatCategories Magers,P. C. 1987 Women, 1986 WeavingatAntelopeHouse.InArchaeologicalInvestiReveal about the Mind. University of Chicago Press, gations at AntelopeHouse, Don P. Morris,SeniorAuthor Chicago. and ProjectDirector,pp. 224-276. NationalParkService, 1991 Cognitive versus GenerativeLinguistics: How ComD.C. Washington, mitments InfluenceResults.Languageand Conmmunication Martin,P. S. 11(1/2):53-62. Colorado.AnthropologiIn Metaphor 1936 LowryRuin in Southwestern Theory of Metaphor. 1993 The Contemporary cal Series,Field Museumof NaturalHistoryVol. 23, No. 1. and Thought,edited by AndrewOrtony,pp. 202-251. 2nd Chicago. Edition.Cambridge UniversityPress,Cambridge. Masterson,K. Lakoff,G., andM. Johnson 1975 BasketryRemainsfromJohnsonCanyon,Colorado.In 1980 MetaphorsWeLive By. Universityof Chicago Press, CanyonProject,Reportof InvestiThe1974 Johnson-Lion Chicago. gation I, assembled by Paul R. Nickens, 223-239. Mesa 1999 Philosophyin the Flesh: The EmbodiedMind and its of Anthropology, UniVerdeResearchCenter,Department Basic Books, New York. Thought. Challengeto Western versityof Colorado,Boulder. Larralde, S. in Prehistoric McCracken, G. 1977 Potteiy andTextileDesign Relationships 1987 Clothingas Language:An ObjectLesson in the Study MA Thesis,DepartArizona,1100-1350 A.D. Unpublished of MaterialCulture.In Materof the ExpressiveProperties ment of Anthropology, Universityof Denver. Approachesto Material ial Anthropology:Contemporary Leach,E. editedby B. ReynoldsandM.A. Stott,pp. 103-128. Culture, 1970 Claude Levi-Strauss. University of Chicago Press, UniversityPress of America,Lanham,MD. Chicago. Sym- Miller,D. TheLogic by Which and Communication: 1976 Culture 1985 Artifactsas Categories.CambridgeUniversityPress, bols Are Connected. CambridgeUniversity Press, CamCambridge. bridge. Morris,E. A. Leh, L. L. Caves in the PrayerRockDistrict,North1980 Basketmaker Reporton the MonumentRuins in San 1942 A Preliminzary easternArizona.Anthropological Papersof the University Juan County,Utah.Universityof ColoradoStudies,Series of Arizona,Number35. Universityof ArizonaPress,TucC, Vol. 1, No. 3. Boulder,CO. son. C. L6vi-Strauss, of ChicagoPress,Chicago. Morris,E. H. 1966 TheSavageMind.University 1927 TheBeginningsof PotteryMakingin theSanJuanArea: Levinson,S. C. UnfiredPrototypesand the Waresof the Earliest Ceramic 1996 Relativity in Spatial Conception and Description. In Period.Anthropological Papersof the AmericanMuseum RethinkingLinguisticRelativity,edited by J. J. Gumperz of NaturalHistory28:125-198. and S. C. Levinson, pp. 177-202. CambridgeUniversity in theLaPlataDistrict.Carnegie 1939 ArchaeologicalStudies Press,Cambridge. D.C. Instituteof Washington Publication519. Washington, L. L6vy-Bruhl, Morris,E. H., andR. F. Burgh 1926 How NativesThink.Knopf,New York. II throughPueblo III. 1941 Anasazi Basketry,Basketnmaker Lightfoot,R. R. Publication 533. Washof Washington Pueblos.In The Processesin Prehistoric CarnegieInstitution 1993 Abandonment and Regions:Ethnoarchaeoington,D.C. Abandonment of Settlemenzts logical and ArchaeologicalApproaches,edited by C. M. Morris,J. N. Uni1991 ArchaeologicalExcavationson the HovenweepLaterCameronand S. A. Tomka,pp. 165-177. Cambridge Number16. als. FourCornersArchaeological Report Project versityPress,Cambridge. CompleteArchaeologicalServiceAssociates,Cortez,CO. Lindsay,L. W., andJ. L. Dykman 1978 Excavation of WestwaterRuin (42SA14), San Juan Neiman,F. D. 1995 Stylistic Variationin EvolutionaryPerspective:InferUtah,FirstFieldSeason.UtahNavajoDevelopment County, Disences from DecorativeDiversityand Interassemblage Council,Blanding,Utah. American Ceramic Assemblages. tancein IllinoisWoodland Lipe, W. D. 60:7-36. Antiquity 1960 1958 Excavations, Glen CanyonAr-ea.University of G. Nordenskiold, UtahAnthropological Papers,Vol. 44. Salt LakeCity. Southwestern Kivas.In TheArchiof Mesa Ver-de, 1990[1893] The CliffDvvellers 1989 Social Scaleof MesaVerdeAnasazi D. LloydMorgan, andImplements. in PrehistoricPueblos, edited Colorado:TheirPottery tectureof Social Integration trans. P.A. Norstedt and Soner, Stockholm and Chicago. Papers byW.D. LipeandM. Hegmon,pp.53-72. Occasional

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

644

AMERICAN ANTIQUITY

[Vol. 65, No. 4, 2000

bridge. 1990 Facsimileedition,Mesa VerdeMuseumAssociation, Renfrew,C., andE. B. W. Zubrow(editors) Mesa Verde. of CognitiveArchaceology. Elemnents 1994 TheAAncientMind: S. G. Ortman, UniversityPress,Cambridge. Cambridge 1998 Corn Grinding and CommunityOrganizationin the andReor- Robertson,I. G. A.D. 1150-1550. InMigration PuebloSouthwest, Analysis, RandomSampling 1999 Spatialand Multivariate Southwest, ganization:thePuebloIVPeriodin theAmerican Error,and AnalyticalNoise: EmpiricalBayesian Methods pp. 165-192. ArizonaStateUnieditedby K. A. Spielmann, 64:137-152. Anitiquity Mexico.Americanl at Teotihuacan, ResearchPapers51, Tempe. versityAnthropological 1999 Using Cognitive Semantics to Discover Continuities Rohn,A. H. 1971 MlugHouse, Mesa VerdeNaltionalPark, Coloracdo. Between Languageand MaterialCulture.Methodsand an ResearchSeries7D. NationalParkService, Archaeological TranExampleUsing Tewaandthe Mesa Verde"Anasazi." D.C. Washington, PeoAffiliationConferenceon Ancestral scriptsandPapers, 1977 Cultural Chanigeanid Continuity oni Chopin Mesa. ples of the FourCornersRegion, vol. 1, pp. 110-120. Fort Regents'Press of Kansas,Lawrence. Lewis College, Durango.CO. 2000 Castle Rock Pueblo Artifacts.In TheArchaeologyof Rosch, E. ansd Ccatego1978 Principlesof Classification.In Cognlitioni CenturyVillagein SouthCastleRockPueblo, a Thirteenth rization,editedby ElanorRoschandB. B. Lloyd,pp. 27-48. western Colorado, edited by K. A. Kuckelman.[HTML Associates,Hillsdale,NJ. LawrenceErlbaum Title]. Available:http://www.crowcanyon.org/researchreandP.BoyesRosch,E., C. B. Mervis,W.D. Gray,D. M. Johnson, ports. Braem S. G., andR. H. Wilshusen Ortman, 1976 Basic Objects in NaturalCategories. Cognitive Psy1996 Letter Report of Salvage-Based Research by Crow chology 8:382-439. CanyonArchaeologicalCenterat the Hedley Ruins, with Recommendationsfor Site Managementand Protection. Salmon,E. Plant 1999 Sharing Breath with Our Relatives: Raranmuri Cortez. Center, Archaeological Reporton file, CrowCanyon Cognitiona Knowledge, Lexicon,anld (Mexico).Unpublished Palmer,G. B. Arizona Ph.D. dissertation.Departmentof Anthropology, a Theoryof CulturalLinguistics.Universityof 1996 Toward StateUniversity. TexasPress,Austin. Sapir,E. Pearson,M., andC. Richards Parker American Culture: inAboriginal Space 1949[1916] TimePerspective of Architecture, Perceptions the World: 1994 Ordering Scapir, of Edvvard A Study in Method.In Selected Writinigs and Order: Approachesto Social andTime.InArchitecture pp. 389-462. Universityof edited by David Mandelbaum, Space, edited by M. ParkerPearsonand C. Richards,pp. CaliforniaPress,Berkeley. 1-37. Routledge,LondonandNew York. reconA Course ofLectutres, 1994 ThePsychologyof Culture: Pepper,G. S. structedandeditedby JudithT. Irvine.Moutonde Gruyter, 1996 PuebloBonito.Universityof New Mexico Press,AlbuBerlin. Papers published1920,Anthropological Originally querque. Vol. 27, AmericanMuseumof NaturalHistory,New York. Schlanger,S. H., andR. H. Wilshusen and Regional Conditionsin the 1993 Local Abandonments Pinker,S. of SettleNorthAmericanSouthwest.In TheAbanidonmenit New York. 1994 TheLanguageInstinct.HarperPerennial, and Archaeologmentsand Regions:Ethnoarchaeological Pope, M. andS. A. Tomka, editedby C. M. Cameron icalApproaches, Thamesand Hudson,New 1999 TheStoryof Decipherment. UniversityPress,Cambridge. pp. 85-98. Cambridge York. A. 0. Shepard, PrestonBlier, S. of 1963 Ceramics jor theArchaeologist.CarnegieInstitution Ontologyand Metaphor 1987 TheAnatomyof Architecture: D.C. Publication609. Washington, Washington in BatammalibaArchitecturalExpression. University of Sillar,B. ChicagoPress,Chicago. 1996 The Dead and the Drying.Jourlnal qf MaterialCulture Reddy,M. J. 1:259-289. A Caseof FrameConflictin Our Metaphor: 1993 TheConduit edited Shore,B. InMetaphor andThought, aboutLanguage. Language 1996 Culturein Mind: Cognitiot, Culture,and the Problem University by A. Ortony,pp. 164-201. 2nd ed. Cambridge OxfordUniversityPress,Oxford. of Meaninig. Press,Cambridge. Sweetser,E. Regier,T. andlCulMetaphor-ical to Pracgmtiatics: 1990 FromEtymology 1996 TheHumanSemanticPotential:SpatialLanguageand University Cambridge turalAspectsof SemanticStructucre. MA. MIT Press,Cambridge, Connectionism. Constrained Press,Cambridge. Renfrew,C. In TheAncientMind: Swentzell,R. aCognitiveArchaeology. 1994 Towards InPuebloStyleanid 1990 PuebloSpace,Form,andMythology. Elementsof CognitiveArchaeology,edited by C. Renfrew W. F E. editedby N. C. Markovich, RegionalArchitectutre, Press, University andE. B. W.Zubrow, pp.3-12. Cambridge Reinhold, andF.G. Sturm,pp.23-30. VanNostrand Preiser, Cambridge. New York. CognitiveArchaeologyandExternal 1998 Mind andMatter: Symbolic Storage.In Cognitionand MaterialCulture:The Teltser,P.A. Theory,and Frequency 1995 CultureHistory,Evolutionary Archaeology of SymbolicStorage,editedby C. Renfrewand Seriation. In EvolutionaryArchaeology:Methoclological CamMonographs, C. Scarre,pp. 1-6. McDonaldInstitute A. Teltser, pp.51-68. Universityof Issues,editedby Patrice bridge. ArizonaPress,Tucson. Renfrew,C., andC. Scarre(editors) B. Loosle, C., J. R. Allison, S. A. Baker,J. C. Janetski, 1998 Cognitionand Material Culture:The Archaeologyof Thompson, andJ. D. Wilde CamSymbolicStorage.McDonaldInstituteMonographs,

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

Ortman]

CONCEPTUAL METAPHOR IN THE ARCHAEOLOGICAL RECORD

645

1988 The Nancy Patterson Village ArchaeologicalResearch Project, Field Year 1986-Preliminary Report No. 4. Museum of Peoples and Cultures Technical Series No. 87-24. BrighamYoungUniversity,Provo. Tilley,C. 1999 Metaphorand MaterialCulture. Blackwell,London. Tootell,R. B. H., E. Switkes,M. S. Silverman,and S. L. Hamilton 1988 FunctionalAnatomy of Macaque Striate Cortex. II. Retinotopic organization. The Journal of Neuroscience
8:1531-1568. Turner, T.

1991 "WeAre Parrots," "Twins Are Birds":Play of Tropesas In BeyondMetaphor:TheTheoryof Structure. Operational Tropesin Anthropology,edited by J. W. Fernandez,pp. 121-158. StanfordUniversityPress, Stanford, CA. Varela,F. J., E. Thompson,andE. Rosch 1991 The EmbodiedMind: CognitiveScience and Human Experience.MIT Press,Cambridge, MA. Varien,M. D. 1999 Sedentismand Mobilityin a Social Landscape:Mesa Verde and Beyond.Universityof ArizonaPress,Tucson. Varien,M. D., W. D. Lipe, M. A. Adler,I. M. Thompson,andB. A. Bradley 1996 Southwest Colorado andSoutheast PatUtahSettlement ternsA.D. 1100 to 1300. In ThePrehistoricPueblo World, A.D. 1150-1350, edited by MichaelA. Adler,pp. 86-113. Universityof ArizonaPress,Tucson. Walens,S. 1981 Feasting with Cannibals:An Essay on KwakiutlCosNJ. mology.PrincetonUniversityPress,Princeton, J. T. Walker, 1977 ML 1147:The Salvageof an Undisturbed Archaeological Site, Manti-Lasal NationalForest,Utah. Unpublished MA Thesis, Department of Anthropology, BrighamYoung University,Provo. Wilshusen,R. H., M. J. Churchill,andJ. M. Potter 1997 PrehistoricReservoirsand WaterBasins in the Mesa VerdeRegion:Intensification of Water CollectionStrategies during the Great Pueblo Period. American Antiquity 62:664-682. Windes,T. C., andD. Ford TheChronometric 1996 TheChacoWood Project: Reappraisal of PuebloBonito.American 61:295-310. Antiquity

Notes

1. As examples: eating and vomiting were primary metaphorsof transformation among the historic Kwakiutl of British Columbia (Walens 1981), and these concepts were expressed in the terminologyof the corporatehousehold and the potlatch,the groundingof animal symbolism in their oral characteristics,the decoration of house doorways as animal of WestAfrica conceptualmouths,etc. Also, the Batammaliba ize the cosmos as a house, and the house as a person, and these metaphorsare expressedin house and body partterminologies, origin myths, rituals associated with house construction,and the physical appearance of houses (PrestonBlier 1987). 2. Examples of psychological results relatingto metaphor include:people automaticallyinfer the meaningsof metaphorical expressionseven when they are not aware of the fact that Received September23, 1998; RevisedDecember 20, 1999; the understood utterance is metaphorical;they comprehend Accepted January 12, 2000.

metaphorical statements just as rapidlyas literalones; and they recall details of written passages more clearly when they are For reviewsof psychologicalstudsummarized metaphorically. ies of metaphor,see Gibbs (1994) and Lakoff and Johnson (1999). 3. In the case study,I use the term "MesaVerde"to referto the Mesa Verderegion in Figure 1, unless otherwisenoted. 4. In calculatinganalogousfeaturefrequencies,only local, whitewaresherds, and the largest sherd of each group from a single vessel, have been considered.If the presenceor absence of a given analogous feature was indeterminateon a given sherd, that sherd did not contributeto the frequency calculations for thatfeature. 5. Complex framingpatternswere defined as patternsthat containmore thanone continuous,detachedline runningparallel to the rim above the interiordesign of a potterybowl. A single thick detachedline is not consideredto representa complex framingpatternbecause a single line does not convey the texture of a coiled basket, and also because the rim of a plaited basketis texturallydistinct,and could have been naturallyrepresentedin potteryusing a single thick line at the rim. 6. Robertson(1999) provides extensive discussion of and justification for the use of empirical Bayesian estimation for spatial samples of varying size. I used a two-stage procedure modeled after Robertson's analysis of TeotihuacanMapping Projectdata. The first stage evaluatedraw proportions of analogousfeaturesin light of the global distribution of proportions acrossthe 11 localities. The following parameters were calculatedfor use in Robertson's equations: For each analogous feature within each locality, n = the numberof sherds on which the presence x= or absence of the featurein question could be determined; the numberof sherdson which the featureoccurs;p = x/n; ,u'= mean of p values for each featureacross all localities;and 2 = g varianceof p values for each featureacross all localities. Using these parameters, a and b values for the priorbeta distribution were calculatedusing Robertson'sequations4a and 4b. I used these values to calculate the most probablevalue, or mode, of the posteriorprobabilitydensity function using the equationin Robertson'snote 2. The mode was chosen insteadof the mean, as given in Robertson's equation 5a, following Robertson's (1999: note 2) recommendation. The second stage evaluatedposteriorestimatesfor each feaof estimates in ture in each locality in light of the distribution local spatial neighborhoods.In this stage, for each analogous featurein each locality, n and x were unchanged;p = mode of the posteriorprobabilitydensity functionas calculatedin stage 1; ,u'= mean p for a local subgroupthatincludesthe locality in question and its two nearest neighbors, all weighted equally regardlessof distance or sample size; and 62 = variance for these same subgroups.Using these parameters,a, b, and the posteriormode were calculated again as in stage 1. This secondaryposteriormode was used in the subsequentPCA.

This content downloaded from 193.54.110.35 on Tue, 4 Mar 2014 07:23:05 AM All use subject to JSTOR Terms and Conditions

S-ar putea să vă placă și